diff --git a/packages/twenty-front/src/modules/ai/components/AIChatCreditsExhaustedMessage.tsx b/packages/twenty-front/src/modules/ai/components/AIChatCreditsExhaustedMessage.tsx
index fa023b328ab2d..c37e771fc2bb3 100644
--- a/packages/twenty-front/src/modules/ai/components/AIChatCreditsExhaustedMessage.tsx
+++ b/packages/twenty-front/src/modules/ai/components/AIChatCreditsExhaustedMessage.tsx
@@ -1,67 +1,28 @@
 import { AIChatBanner } from '@/ai/components/AIChatBanner';
-import { useEndSubscriptionTrialPeriod } from '@/settings/billing/hooks/useEndSubscriptionTrialPeriod';
-import { useRedirect } from '@/domain-manager/hooks/useRedirect';
 import { usePermissionFlagMap } from '@/settings/roles/hooks/usePermissionFlagMap';
 import { useSubscriptionStatus } from '@/workspace/hooks/useSubscriptionStatus';
 import { t } from '@lingui/core/macro';
-import { useState } from 'react';
 import { SettingsPath } from 'twenty-shared/types';
-import { getSettingsPath, isDefined } from 'twenty-shared/utils';
 import { IconSparkles } from 'twenty-ui/display';
-import { useQuery } from '@apollo/client/react';
 import {
   PermissionFlagType,
   SubscriptionStatus,
-  BillingPortalSessionDocument,
 } from '~/generated-metadata/graphql';
+import { useNavigateSettings } from '~/hooks/useNavigateSettings';
 
 export const AIChatCreditsExhaustedMessage = () => {
-  const { redirect } = useRedirect();
+  const navigateSettings = useNavigateSettings();
   const subscriptionStatus = useSubscriptionStatus();
-  const { endTrialPeriod, isLoading: isEndingTrial } =
-    useEndSubscriptionTrialPeriod();
-  const [isProcessing, setIsProcessing] = useState(false);
 
   const isTrialing = subscriptionStatus === SubscriptionStatus.Trialing;
 
   const { [PermissionFlagType.WORKSPACE]: hasPermissionToManageBilling } =
     usePermissionFlagMap();
 
-  const { data: billingPortalData, loading: isBillingPortalLoading } = useQuery(
-    BillingPortalSessionDocument,
-    {
-      variables: {
-        returnUrlPath: getSettingsPath(SettingsPath.Billing),
-      },
-    },
-  );
-
-  const openBillingPortal = () => {
-    if (
-      isDefined(billingPortalData) &&
-      isDefined(billingPortalData.billingPortalSession.url)
-    ) {
-      redirect(billingPortalData.billingPortalSession.url);
-    }
+  const handleUpgradeClick = () => {
+    navigateSettings(SettingsPath.Billing);
   };
 
-  const handleUpgradeClick = async () => {
-    if (!isTrialing) {
-      openBillingPortal();
-      return;
-    }
-
-    setIsProcessing(true);
-    const result = await endTrialPeriod();
-    setIsProcessing(false);
-
-    if (!result.success) {
-      openBillingPortal();
-    }
-  };
-
-  const isLoading = isEndingTrial || isBillingPortalLoading || isProcessing;
-
   const message = hasPermissionToManageBilling
     ? isTrialing
       ? t`Free trial credits exhausted. Subscribe now to continue using AI features.`
@@ -79,8 +40,6 @@ export const AIChatCreditsExhaustedMessage = () => {
       buttonOnClick={
         hasPermissionToManageBilling ? handleUpgradeClick : undefined
       }
-      isButtonDisabled={isLoading}
-      isButtonLoading={isLoading}
     />
   );
 };
diff --git a/packages/twenty-front/src/modules/ai/components/AIChatEditorSection.tsx b/packages/twenty-front/src/modules/ai/components/AIChatEditorSection.tsx
index c04638e3bdef4..04667ca3f057b 100644
--- a/packages/twenty-front/src/modules/ai/components/AIChatEditorSection.tsx
+++ b/packages/twenty-front/src/modules/ai/components/AIChatEditorSection.tsx
@@ -1,6 +1,5 @@
 import { styled } from '@linaria/react';
 import { EditorContent } from '@tiptap/react';
-import { LightButton } from 'twenty-ui/input';
 import { themeCssVariables } from 'twenty-ui/theme-constants';
 
 import { AIChatEmptyState } from '@/ai/components/AIChatEmptyState';
@@ -12,10 +11,17 @@ import { AIChatEditorFocusEffect } from '@/ai/components/internal/AIChatEditorFo
 import { AIChatSkeletonLoader } from '@/ai/components/internal/AIChatSkeletonLoader';
 import { SendMessageButton } from '@/ai/components/internal/SendMessageButton';
 import { useAIChatEditor } from '@/ai/hooks/useAIChatEditor';
-import { useAiModelLabel } from '@/ai/hooks/useAiModelOptions';
+import { useAgentChatModelId } from '@/ai/hooks/useAgentChatModelId';
+import { useWorkspaceAiModelAvailability } from '@/ai/hooks/useWorkspaceAiModelAvailability';
+import { agentChatUserSelectedModelState } from '@/ai/states/agentChatUserSelectedModelState';
 import { currentWorkspaceState } from '@/auth/states/currentWorkspaceState';
+import { aiModelsState } from '@/client-config/states/aiModelsState';
+import { getModelIcon } from '@/settings/admin-panel/ai/utils/getModelIcon';
+import { Select } from '@/ui/input/components/Select';
 import { useIsMobile } from '@/ui/utilities/responsive/hooks/useIsMobile';
+import { useSetAtomState } from '@/ui/utilities/state/jotai/hooks/useSetAtomState';
 import { useAtomStateValue } from '@/ui/utilities/state/jotai/hooks/useAtomStateValue';
+import { t } from '@lingui/core/macro';
 
 const StyledInputArea = styled.div<{ isMobile: boolean }>`
   align-items: flex-end;
@@ -102,24 +108,48 @@ const StyledRightButtonsContainer = styled.div`
   gap: ${themeCssVariables.spacing[1]};
 `;
 
-const StyledReadOnlyModelButtonContainer = styled.div`
-  > * {
-    cursor: default;
-
-    &:hover,
-    &:active {
-      background: transparent;
-    }
-  }
-`;
-
 export const AIChatEditorSection = () => {
   const isMobile = useIsMobile();
   const currentWorkspace = useAtomStateValue(currentWorkspaceState);
-  const smartModelLabel = useAiModelLabel(currentWorkspace?.smartModel, false);
+  const aiModels = useAtomStateValue(aiModelsState);
+  const { enabledModels } = useWorkspaceAiModelAvailability();
+  const setAgentChatUserSelectedModel = useSetAtomState(
+    agentChatUserSelectedModelState,
+  );
+  const { selectedModelId } = useAgentChatModelId();
 
   const { editor, handleSendAndClear } = useAIChatEditor();
 
+  const workspaceSmartModel = aiModels.find(
+    (model) => model.modelId === currentWorkspace?.smartModel,
+  );
+
+  const resolvedDefaultModelId = enabledModels.find(
+    (model) =>
+      model.label === workspaceSmartModel?.label &&
+      model.providerName === workspaceSmartModel?.providerName,
+  )?.modelId;
+
+  const defaultPinnedOption = workspaceSmartModel
+    ? {
+        value: null as string | null,
+        label: workspaceSmartModel.label,
+        Icon: getModelIcon(
+          workspaceSmartModel.modelFamily,
+          workspaceSmartModel.providerName,
+        ),
+        contextualText: t`default`,
+      }
+    : undefined;
+
+  const smartModelOptions = enabledModels
+    .filter((model) => model.modelId !== resolvedDefaultModelId)
+    .map((model) => ({
+      value: model.modelId as string | null,
+      label: model.label,
+      Icon: getModelIcon(model.modelFamily, model.providerName),
+    }));
+
   return (
     <>
       <AIChatEditorFocusEffect editor={editor} />
@@ -139,9 +169,16 @@ export const AIChatEditorSection = () => {
               <AIChatContextUsageButton />
             </StyledLeftButtonsContainer>
             <StyledRightButtonsContainer>
-              <StyledReadOnlyModelButtonContainer>
-                <LightButton accent="tertiary" title={smartModelLabel} />
-              </StyledReadOnlyModelButtonContainer>
+              <Select
+                dropdownId="ai-chat-smart-model-select"
+                value={selectedModelId}
+                onChange={setAgentChatUserSelectedModel}
+                options={smartModelOptions}
+                pinnedOption={defaultPinnedOption}
+                selectSizeVariant="small"
+                withSearchInput
+                dropdownOffset={{ x: 0, y: 8 }}
+              />
               <SendMessageButton onSend={handleSendAndClear} />
             </StyledRightButtonsContainer>
           </StyledButtonsContainer>
diff --git a/packages/twenty-front/src/modules/ai/components/AIChatErrorUnderMessageList.tsx b/packages/twenty-front/src/modules/ai/components/AIChatErrorUnderMessageList.tsx
index 1e17d8a8334a5..4e9f61f2ce072 100644
--- a/packages/twenty-front/src/modules/ai/components/AIChatErrorUnderMessageList.tsx
+++ b/packages/twenty-front/src/modules/ai/components/AIChatErrorUnderMessageList.tsx
@@ -1,4 +1,4 @@
-import { AIChatStandaloneError } from '@/ai/components/AIChatStandaloneError';
+import { AIChatErrorRenderer } from '@/ai/components/AIChatErrorRenderer';
 import { AgentMessageRole } from '@/ai/constants/AgentMessageRole';
 import { agentChatErrorState } from '@/ai/states/agentChatErrorState';
 import { agentChatIsStreamingState } from '@/ai/states/agentChatIsStreamingState';
@@ -7,6 +7,12 @@ import { agentChatMessageIdsComponentSelector } from '@/ai/states/agentChatMessa
 import { useAtomComponentFamilySelectorValue } from '@/ui/utilities/state/jotai/hooks/useAtomComponentFamilySelectorValue';
 import { useAtomComponentSelectorValue } from '@/ui/utilities/state/jotai/hooks/useAtomComponentSelectorValue';
 import { useAtomStateValue } from '@/ui/utilities/state/jotai/hooks/useAtomStateValue';
+import { styled } from '@linaria/react';
+import { themeCssVariables } from 'twenty-ui/theme-constants';
+
+const StyledErrorWrapper = styled.div`
+  padding-top: ${themeCssVariables.spacing[3]};
+`;
 
 export const AIChatErrorUnderMessageList = () => {
   const agentChatError = useAtomStateValue(agentChatErrorState);
@@ -31,5 +37,9 @@ export const AIChatErrorUnderMessageList = () => {
     return null;
   }
 
-  return <AIChatStandaloneError />;
+  return (
+    <StyledErrorWrapper>
+      <AIChatErrorRenderer error={agentChatError} />
+    </StyledErrorWrapper>
+  );
 };
diff --git a/packages/twenty-front/src/modules/ai/components/AgentChatAiSdkStreamEffect.tsx b/packages/twenty-front/src/modules/ai/components/AgentChatAiSdkStreamEffect.tsx
index b728822d6dc21..e0df67e2e551d 100644
--- a/packages/twenty-front/src/modules/ai/components/AgentChatAiSdkStreamEffect.tsx
+++ b/packages/twenty-front/src/modules/ai/components/AgentChatAiSdkStreamEffect.tsx
@@ -6,6 +6,7 @@ import { useEnsureAgentChatThreadIdForSend } from '@/ai/hooks/useEnsureAgentChat
 import { agentChatErrorState } from '@/ai/states/agentChatErrorState';
 import { agentChatIsLoadingState } from '@/ai/states/agentChatIsLoadingState';
 import { agentChatIsStreamingState } from '@/ai/states/agentChatIsStreamingState';
+import { normalizeAiSdkError } from '@/ai/utils/normalizeAiSdkError';
 import { agentChatMessagesComponentFamilyState } from '@/ai/states/agentChatMessagesComponentFamilyState';
 import { agentChatMessagesLoadingState } from '@/ai/states/agentChatMessagesLoadingState';
 import { agentChatThreadsLoadingState } from '@/ai/states/agentChatThreadsLoadingState';
@@ -119,7 +120,7 @@ export const AgentChatAiSdkStreamEffect = () => {
   const setAgentChatError = useSetAtomState(agentChatErrorState);
 
   useEffect(() => {
-    setAgentChatError(chatState.error);
+    setAgentChatError(normalizeAiSdkError(chatState.error));
   }, [chatState.error, setAgentChatError]);
 
   const setAgentChatIsStreaming = useSetAtomState(agentChatIsStreamingState);
diff --git a/packages/twenty-front/src/modules/ai/constants/DefaultFastModel.ts b/packages/twenty-front/src/modules/ai/constants/DefaultFastModel.ts
deleted file mode 100644
index 84e98b83087b1..0000000000000
--- a/packages/twenty-front/src/modules/ai/constants/DefaultFastModel.ts
+++ /dev/null
@@ -1 +0,0 @@
-export const DEFAULT_FAST_MODEL = 'default-fast-model' as const;
diff --git a/packages/twenty-front/src/modules/ai/constants/DefaultSmartModel.ts b/packages/twenty-front/src/modules/ai/constants/DefaultSmartModel.ts
deleted file mode 100644
index f85d6f971916b..0000000000000
--- a/packages/twenty-front/src/modules/ai/constants/DefaultSmartModel.ts
+++ /dev/null
@@ -1 +0,0 @@
-export const DEFAULT_SMART_MODEL = 'default-smart-model' as const;
diff --git a/packages/twenty-front/src/modules/ai/hooks/useAgentChat.ts b/packages/twenty-front/src/modules/ai/hooks/useAgentChat.ts
index 9bd1d07701830..393f176a5a4a6 100644
--- a/packages/twenty-front/src/modules/ai/hooks/useAgentChat.ts
+++ b/packages/twenty-front/src/modules/ai/hooks/useAgentChat.ts
@@ -15,6 +15,7 @@ import {
   agentChatDraftsByThreadIdState,
 } from '@/ai/states/agentChatDraftsByThreadIdState';
 import { agentChatInputState } from '@/ai/states/agentChatInputState';
+import { useAgentChatModelId } from '@/ai/hooks/useAgentChatModelId';
 import { REST_API_BASE_URL } from '@/apollo/constant/rest-api-base-url';
 import { getTokenPair } from '@/apollo/utils/getTokenPair';
 import { renewToken } from '@/auth/services/AuthService';
@@ -40,6 +41,7 @@ export const useAgentChat = (
   const setTokenPair = useSetAtomState(tokenPairState);
   const setAgentChatUsage = useSetAtomState(agentChatUsageState);
 
+  const { modelIdForRequest } = useAgentChatModelId();
   const { getBrowsingContext } = useGetBrowsingContext();
   const setCurrentAIChatThreadTitle = useSetAtomState(
     currentAIChatThreadTitleState,
@@ -258,6 +260,9 @@ export const useAgentChat = (
         body: {
           threadId,
           browsingContext,
+          ...(isDefined(modelIdForRequest) && {
+            modelId: modelIdForRequest,
+          }),
         },
       },
     );
@@ -273,6 +278,7 @@ export const useAgentChat = (
     agentChatUploadedFiles,
     setAgentChatUploadedFiles,
     setAgentChatDraftsByThreadId,
+    modelIdForRequest,
   ]);
 
   useListenToBrowserEvent({
diff --git a/packages/twenty-front/src/modules/ai/hooks/useAgentChatModelId.ts b/packages/twenty-front/src/modules/ai/hooks/useAgentChatModelId.ts
new file mode 100644
index 0000000000000..661965d63cf66
--- /dev/null
+++ b/packages/twenty-front/src/modules/ai/hooks/useAgentChatModelId.ts
@@ -0,0 +1,23 @@
+import { isDefined } from 'twenty-shared/utils';
+
+import { useWorkspaceAiModelAvailability } from '@/ai/hooks/useWorkspaceAiModelAvailability';
+import { agentChatUserSelectedModelState } from '@/ai/states/agentChatUserSelectedModelState';
+import { useAtomStateValue } from '@/ui/utilities/state/jotai/hooks/useAtomStateValue';
+
+export const useAgentChatModelId = () => {
+  const { enabledModels } = useWorkspaceAiModelAvailability();
+  const agentChatUserSelectedModel = useAtomStateValue(
+    agentChatUserSelectedModelState,
+  );
+
+  const isUserModelAvailable =
+    !isDefined(agentChatUserSelectedModel) ||
+    enabledModels.some((model) => model.modelId === agentChatUserSelectedModel);
+
+  const selectedModelId = isUserModelAvailable
+    ? agentChatUserSelectedModel
+    : null;
+  const modelIdForRequest = selectedModelId ?? undefined;
+
+  return { selectedModelId, modelIdForRequest };
+};
diff --git a/packages/twenty-front/src/modules/ai/hooks/useAiModelOptions.ts b/packages/twenty-front/src/modules/ai/hooks/useAiModelOptions.ts
index dd902fba3decc..27398e3543b64 100644
--- a/packages/twenty-front/src/modules/ai/hooks/useAiModelOptions.ts
+++ b/packages/twenty-front/src/modules/ai/hooks/useAiModelOptions.ts
@@ -1,7 +1,6 @@
 import { type SelectOption } from 'twenty-ui/input';
+import { isAutoSelectModelId } from 'twenty-shared/utils';
 
-import { DEFAULT_FAST_MODEL } from '@/ai/constants/DefaultFastModel';
-import { DEFAULT_SMART_MODEL } from '@/ai/constants/DefaultSmartModel';
 import { useWorkspaceAiModelAvailability } from '@/ai/hooks/useWorkspaceAiModelAvailability';
 import { aiModelsState } from '@/client-config/states/aiModelsState';
 import { useAtomStateValue } from '@/ui/utilities/state/jotai/hooks/useAtomStateValue';
@@ -16,13 +15,11 @@ export const useAiModelOptions = (): SelectOption<string>[] => {
     )
     .map((model) => ({
       value: model.modelId,
-      label:
-        model.modelId === DEFAULT_FAST_MODEL ||
-        model.modelId === DEFAULT_SMART_MODEL
-          ? model.label
-          : model.modelFamilyLabel
-            ? `${model.label} (${model.modelFamilyLabel})`
-            : model.label,
+      label: isAutoSelectModelId(model.modelId)
+        ? model.label
+        : model.modelFamilyLabel
+          ? `${model.label} (${model.modelFamilyLabel})`
+          : model.label,
     }))
     .sort((a, b) => a.label.localeCompare(b.label));
 };
@@ -43,11 +40,7 @@ export const useAiModelLabel = (
     return modelId;
   }
 
-  if (
-    model.modelId === DEFAULT_FAST_MODEL ||
-    model.modelId === DEFAULT_SMART_MODEL ||
-    !includeProvider
-  ) {
+  if (isAutoSelectModelId(model.modelId) || !includeProvider) {
     return model.label;
   }
 
diff --git a/packages/twenty-front/src/modules/ai/hooks/useWorkspaceAiModelAvailability.ts b/packages/twenty-front/src/modules/ai/hooks/useWorkspaceAiModelAvailability.ts
index 4680a1567ce98..7e844289d8afa 100644
--- a/packages/twenty-front/src/modules/ai/hooks/useWorkspaceAiModelAvailability.ts
+++ b/packages/twenty-front/src/modules/ai/hooks/useWorkspaceAiModelAvailability.ts
@@ -1,17 +1,10 @@
-import { DEFAULT_FAST_MODEL } from '@/ai/constants/DefaultFastModel';
-import { DEFAULT_SMART_MODEL } from '@/ai/constants/DefaultSmartModel';
+import { isAutoSelectModelId } from 'twenty-shared/utils';
+
 import { currentWorkspaceState } from '@/auth/states/currentWorkspaceState';
 import { aiModelsState } from '@/client-config/states/aiModelsState';
 import { useAtomStateValue } from '@/ui/utilities/state/jotai/hooks/useAtomStateValue';
 import { type ClientAiModelConfig } from '~/generated-metadata/graphql';
 
-const VIRTUAL_MODEL_IDS: Set<string> = new Set([
-  DEFAULT_SMART_MODEL,
-  DEFAULT_FAST_MODEL,
-]);
-
-const isVirtualModel = (modelId: string) => VIRTUAL_MODEL_IDS.has(modelId);
-
 export const useWorkspaceAiModelAvailability = () => {
   const aiModels = useAtomStateValue(aiModelsState);
   const currentWorkspace = useAtomStateValue(currentWorkspaceState);
@@ -23,7 +16,7 @@ export const useWorkspaceAiModelAvailability = () => {
     modelId: string,
     model?: ClientAiModelConfig,
   ): boolean => {
-    if (isVirtualModel(modelId)) {
+    if (isAutoSelectModelId(modelId)) {
       return true;
     }
 
@@ -35,7 +28,7 @@ export const useWorkspaceAiModelAvailability = () => {
   };
 
   const realModels = aiModels.filter(
-    (model) => !isVirtualModel(model.modelId) && !model.isDeprecated,
+    (model) => !isAutoSelectModelId(model.modelId) && !model.isDeprecated,
   );
 
   const enabledModels = realModels.filter((model) =>
diff --git a/packages/twenty-front/src/modules/ai/states/agentChatUserSelectedModelState.ts b/packages/twenty-front/src/modules/ai/states/agentChatUserSelectedModelState.ts
new file mode 100644
index 0000000000000..0383c09b14ec8
--- /dev/null
+++ b/packages/twenty-front/src/modules/ai/states/agentChatUserSelectedModelState.ts
@@ -0,0 +1,8 @@
+import { createAtomState } from '@/ui/utilities/state/jotai/utils/createAtomState';
+
+export const agentChatUserSelectedModelState = createAtomState<string | null>({
+  key: 'ai/agentChatUserSelectedModel',
+  defaultValue: null,
+  useLocalStorage: true,
+  localStorageOptions: { getOnInit: true },
+});
diff --git a/packages/twenty-front/src/modules/ai/utils/__tests__/normalizeAiSdkError.test.ts b/packages/twenty-front/src/modules/ai/utils/__tests__/normalizeAiSdkError.test.ts
new file mode 100644
index 0000000000000..50088384fc18e
--- /dev/null
+++ b/packages/twenty-front/src/modules/ai/utils/__tests__/normalizeAiSdkError.test.ts
@@ -0,0 +1,60 @@
+import { normalizeAiSdkError } from '@/ai/utils/normalizeAiSdkError';
+
+describe('normalizeAiSdkError', () => {
+  it('should return undefined for undefined input', () => {
+    expect(normalizeAiSdkError(undefined)).toBeUndefined();
+  });
+
+  it('should return the error unchanged if it already has a code', () => {
+    const error = new Error('test') as Error & { code: string };
+    error.code = 'EXISTING_CODE';
+
+    const result = normalizeAiSdkError(error);
+
+    expect(result).toBe(error);
+    expect((result as Error & { code: string }).code).toBe('EXISTING_CODE');
+  });
+
+  it('should parse JSON message and attach code from response body', () => {
+    const error = new Error(
+      '{"statusCode":402,"error":"Error","messages":["Credits exhausted"],"code":"BILLING_CREDITS_EXHAUSTED"}',
+    );
+
+    const result = normalizeAiSdkError(error);
+
+    expect(result).not.toBe(error);
+    expect((result as Error & { code: string }).code).toBe(
+      'BILLING_CREDITS_EXHAUSTED',
+    );
+    expect(result?.message).toBe(error.message);
+  });
+
+  it('should return original error for non-JSON message', () => {
+    const error = new Error('Something went wrong');
+
+    const result = normalizeAiSdkError(error);
+
+    expect(result).toBe(error);
+  });
+
+  it('should return original error for JSON message without code', () => {
+    const error = new Error(
+      '{"statusCode":500,"error":"Internal Server Error"}',
+    );
+
+    const result = normalizeAiSdkError(error);
+
+    expect(result).toBe(error);
+  });
+
+  it('should preserve the original stack trace', () => {
+    const error = new Error(
+      '{"statusCode":402,"code":"BILLING_CREDITS_EXHAUSTED"}',
+    );
+    const originalStack = error.stack;
+
+    const result = normalizeAiSdkError(error);
+
+    expect(result?.stack).toBe(originalStack);
+  });
+});
diff --git a/packages/twenty-front/src/modules/ai/utils/normalizeAiSdkError.ts b/packages/twenty-front/src/modules/ai/utils/normalizeAiSdkError.ts
new file mode 100644
index 0000000000000..bca90d57f1ee7
--- /dev/null
+++ b/packages/twenty-front/src/modules/ai/utils/normalizeAiSdkError.ts
@@ -0,0 +1,43 @@
+import { isDefined } from 'twenty-shared/utils';
+
+// The Vercel AI SDK wraps non-200 responses into an Error whose message
+// is the raw JSON response body, losing any custom properties (like `code`).
+// This function parses that JSON and re-attaches the code so downstream
+// consumers (extractErrorCode) can find it without knowing about the SDK.
+export const normalizeAiSdkError = (
+  error: Error | undefined,
+): Error | undefined => {
+  if (!isDefined(error) || !(error instanceof Error)) {
+    return error;
+  }
+
+  if (
+    'code' in error &&
+    typeof (error as Error & { code: string }).code === 'string'
+  ) {
+    return error;
+  }
+
+  try {
+    const parsed: unknown = JSON.parse(error.message);
+
+    if (
+      isDefined(parsed) &&
+      typeof parsed === 'object' &&
+      'code' in parsed &&
+      typeof (parsed as { code: unknown }).code === 'string'
+    ) {
+      const normalizedError = new Error(error.message) as Error & {
+        code: string;
+      };
+      normalizedError.code = (parsed as { code: string }).code;
+      normalizedError.stack = error.stack;
+
+      return normalizedError;
+    }
+  } catch {
+    // message is not JSON — nothing to normalize
+  }
+
+  return error;
+};
diff --git a/packages/twenty-front/src/modules/apollo/services/__tests__/apollo.factory.test.ts b/packages/twenty-front/src/modules/apollo/services/__tests__/apollo.factory.test.ts
index 1d7f029277d8f..97f75257e71d8 100644
--- a/packages/twenty-front/src/modules/apollo/services/__tests__/apollo.factory.test.ts
+++ b/packages/twenty-front/src/modules/apollo/services/__tests__/apollo.factory.test.ts
@@ -2,8 +2,10 @@ import { gql, InMemoryCache } from '@apollo/client';
 import { CombinedGraphQLErrors } from '@apollo/client/errors';
 import fetchMock, { enableFetchMocks } from 'jest-fetch-mock';
 
-import { DEFAULT_FAST_MODEL } from '@/ai/constants/DefaultFastModel';
-import { DEFAULT_SMART_MODEL } from '@/ai/constants/DefaultSmartModel';
+import {
+  AUTO_SELECT_FAST_MODEL_ID,
+  AUTO_SELECT_SMART_MODEL_ID,
+} from 'twenty-shared/constants';
 import { ApolloFactory, type Options } from '@/apollo/services/apollo.factory';
 import { CUSTOM_WORKSPACE_APPLICATION_MOCK } from '@/object-metadata/hooks/__tests__/constants/CustomWorkspaceApplicationMock.test.constant';
 import { WorkspaceActivationStatus } from '~/generated-metadata/graphql';
@@ -71,8 +73,8 @@ const mockWorkspace = {
   isTwoFactorAuthenticationEnforced: false,
   trashRetentionDays: 14,
   eventLogRetentionDays: 365 * 3,
-  fastModel: DEFAULT_FAST_MODEL,
-  smartModel: DEFAULT_SMART_MODEL,
+  fastModel: AUTO_SELECT_FAST_MODEL_ID,
+  smartModel: AUTO_SELECT_SMART_MODEL_ID,
   routerModel: 'auto',
   enabledAiModelIds: [],
   useRecommendedModels: true,
diff --git a/packages/twenty-front/src/modules/object-metadata/hooks/__tests__/useColumnDefinitionsFromObjectMetadata.test.ts b/packages/twenty-front/src/modules/object-metadata/hooks/__tests__/useColumnDefinitionsFromObjectMetadata.test.ts
index e77994d0ef08f..2643bc66ee648 100644
--- a/packages/twenty-front/src/modules/object-metadata/hooks/__tests__/useColumnDefinitionsFromObjectMetadata.test.ts
+++ b/packages/twenty-front/src/modules/object-metadata/hooks/__tests__/useColumnDefinitionsFromObjectMetadata.test.ts
@@ -1,7 +1,9 @@
 import { renderHook } from '@testing-library/react';
 
-import { DEFAULT_FAST_MODEL } from '@/ai/constants/DefaultFastModel';
-import { DEFAULT_SMART_MODEL } from '@/ai/constants/DefaultSmartModel';
+import {
+  AUTO_SELECT_FAST_MODEL_ID,
+  AUTO_SELECT_SMART_MODEL_ID,
+} from 'twenty-shared/constants';
 import { currentWorkspaceState } from '@/auth/states/currentWorkspaceState';
 import { CUSTOM_WORKSPACE_APPLICATION_MOCK } from '@/object-metadata/hooks/__tests__/constants/CustomWorkspaceApplicationMock.test.constant';
 import { useColumnDefinitionsFromObjectMetadata } from '@/object-metadata/hooks/useColumnDefinitionsFromObjectMetadata';
@@ -68,8 +70,8 @@ describe('useColumnDefinitionsFromObjectMetadata', () => {
       isTwoFactorAuthenticationEnforced: false,
       trashRetentionDays: 14,
       eventLogRetentionDays: 365 * 3,
-      fastModel: DEFAULT_FAST_MODEL,
-      smartModel: DEFAULT_SMART_MODEL,
+      fastModel: AUTO_SELECT_FAST_MODEL_ID,
+      smartModel: AUTO_SELECT_SMART_MODEL_ID,
       enabledAiModelIds: [],
       useRecommendedModels: true,
     });
diff --git a/packages/twenty-front/src/modules/settings/experience/components/DateTimeSettingsDateFormatSelect.tsx b/packages/twenty-front/src/modules/settings/experience/components/DateTimeSettingsDateFormatSelect.tsx
index fe349367fa3aa..5f2ced128c36f 100644
--- a/packages/twenty-front/src/modules/settings/experience/components/DateTimeSettingsDateFormatSelect.tsx
+++ b/packages/twenty-front/src/modules/settings/experience/components/DateTimeSettingsDateFormatSelect.tsx
@@ -34,38 +34,38 @@ export const DateTimeSettingsDateFormatSelect = ({
   return (
     <Select
       dropdownId="datetime-settings-date-format"
-      dropdownWidth={218}
+      dropdownWidth={320}
       label={t`Date format`}
       fullWidth
-      dropdownWidthAuto
       value={value}
+      pinnedOption={{
+        label: t`System settings`,
+        value: DateFormat.SYSTEM,
+        contextualText: systemDateFormatLabel,
+      }}
       options={[
         {
-          label: t`System settings - ${systemDateFormatLabel}`,
-          value: DateFormat.SYSTEM,
-        },
-        {
-          label: `${formatInTimeZone(
+          label: formatInTimeZone(
             Date.now(),
             usedTimeZone,
             DateFormat.MONTH_FIRST,
-          )}`,
+          ),
           value: DateFormat.MONTH_FIRST,
         },
         {
-          label: `${formatInTimeZone(
+          label: formatInTimeZone(
             Date.now(),
             usedTimeZone,
             DateFormat.DAY_FIRST,
-          )}`,
+          ),
           value: DateFormat.DAY_FIRST,
         },
         {
-          label: `${formatInTimeZone(
+          label: formatInTimeZone(
             Date.now(),
             usedTimeZone,
             DateFormat.YEAR_FIRST,
-          )}`,
+          ),
           value: DateFormat.YEAR_FIRST,
         },
       ]}
diff --git a/packages/twenty-front/src/modules/settings/experience/components/DateTimeSettingsTimeFormatSelect.tsx b/packages/twenty-front/src/modules/settings/experience/components/DateTimeSettingsTimeFormatSelect.tsx
index 084ebc310ee4d..9214496d2b0ef 100644
--- a/packages/twenty-front/src/modules/settings/experience/components/DateTimeSettingsTimeFormatSelect.tsx
+++ b/packages/twenty-front/src/modules/settings/experience/components/DateTimeSettingsTimeFormatSelect.tsx
@@ -50,18 +50,21 @@ export const DateTimeSettingsTimeFormatSelect = ({
       dropdownWidthAuto
       fullWidth
       value={value}
+      pinnedOption={{
+        label: t`System settings`,
+        value: TimeFormat.SYSTEM,
+        contextualText: systemTimeFormatLabel,
+      }}
       options={[
         {
-          label: t`System Settings - ${systemTimeFormatLabel}`,
-          value: TimeFormat.SYSTEM,
-        },
-        {
-          label: t`24h - ${hour24Label}`,
+          label: t`24h`,
           value: TimeFormat.HOUR_24,
+          contextualText: hour24Label,
         },
         {
-          label: t`12h - ${hour12Label}`,
+          label: t`12h`,
           value: TimeFormat.HOUR_12,
+          contextualText: hour12Label,
         },
       ]}
       onChange={onChange}
diff --git a/packages/twenty-front/src/modules/settings/experience/components/DateTimeSettingsTimeZoneSelect.tsx b/packages/twenty-front/src/modules/settings/experience/components/DateTimeSettingsTimeZoneSelect.tsx
index cd9689e001e7b..7089936c7bf77 100644
--- a/packages/twenty-front/src/modules/settings/experience/components/DateTimeSettingsTimeZoneSelect.tsx
+++ b/packages/twenty-front/src/modules/settings/experience/components/DateTimeSettingsTimeZoneSelect.tsx
@@ -3,7 +3,6 @@ import { findAvailableTimeZoneOption } from '@/localization/utils/findAvailableT
 import { AVAILABLE_TIMEZONE_OPTIONS } from '@/settings/experience/constants/AvailableTimezoneOptions';
 import { Select } from '@/ui/input/components/Select';
 import { t } from '@lingui/core/macro';
-import { isDefined } from 'twenty-shared/utils';
 import { type SelectOption } from 'twenty-ui/input';
 
 type DateTimeSettingsTimeZoneSelectProps = {
@@ -23,18 +22,15 @@ export const DateTimeSettingsTimeZoneSelect = ({
     <Select
       dropdownId="datetime-settings-time-zone"
       label={t`Time zone`}
-      dropdownWidthAuto
+      dropdownWidth={480}
       fullWidth
       value={value}
-      options={[
-        {
-          label: isDefined(systemTimeZoneOption)
-            ? t`System settings`.concat(` - ${systemTimeZoneOption.label}`)
-            : t`System settings`,
-          value: 'system',
-        },
-        ...(AVAILABLE_TIMEZONE_OPTIONS as SelectOption<string>[]),
-      ]}
+      pinnedOption={{
+        label: t`System settings`,
+        value: 'system',
+        contextualText: systemTimeZoneOption?.label,
+      }}
+      options={AVAILABLE_TIMEZONE_OPTIONS as SelectOption<string>[]}
       onChange={onChange}
       withSearchInput
     />
diff --git a/packages/twenty-front/src/modules/settings/experience/components/NumberFormatSelect.tsx b/packages/twenty-front/src/modules/settings/experience/components/NumberFormatSelect.tsx
index 66257c2924b57..b5011c8e52b3d 100644
--- a/packages/twenty-front/src/modules/settings/experience/components/NumberFormatSelect.tsx
+++ b/packages/twenty-front/src/modules/settings/experience/components/NumberFormatSelect.tsx
@@ -46,26 +46,31 @@ export const NumberFormatSelect = ({
       dropdownWidthAuto
       fullWidth
       value={value}
+      pinnedOption={{
+        label: t`System settings`,
+        value: NumberFormat.SYSTEM,
+        contextualText: systemNumberFormatLabel,
+      }}
       options={[
         {
-          label: t`System Settings - ${systemNumberFormatLabel}`,
-          value: NumberFormat.SYSTEM,
-        },
-        {
-          label: t`Commas and dot - ${commasAndDotExample}`,
+          label: t`Commas and dot`,
           value: NumberFormat.COMMAS_AND_DOT,
+          contextualText: commasAndDotExample,
         },
         {
-          label: t`Spaces and comma - ${spacesAndCommaExample}`,
+          label: t`Spaces and comma`,
           value: NumberFormat.SPACES_AND_COMMA,
+          contextualText: spacesAndCommaExample,
         },
         {
-          label: t`Dots and comma - ${dotsAndCommaExample}`,
+          label: t`Dots and comma`,
           value: NumberFormat.DOTS_AND_COMMA,
+          contextualText: dotsAndCommaExample,
         },
         {
-          label: t`Apostrophe and dot - ${apostropheAndDotExample}`,
+          label: t`Apostrophe and dot`,
           value: NumberFormat.APOSTROPHE_AND_DOT,
+          contextualText: apostropheAndDotExample,
         },
       ]}
       onChange={onChange}
diff --git a/packages/twenty-front/src/modules/ui/input/components/Select.tsx b/packages/twenty-front/src/modules/ui/input/components/Select.tsx
index 715da8a1b74a8..8e7f79d899bec 100644
--- a/packages/twenty-front/src/modules/ui/input/components/Select.tsx
+++ b/packages/twenty-front/src/modules/ui/input/components/Select.tsx
@@ -49,6 +49,7 @@ export type SelectProps<Value extends SelectValue> = {
   value?: Value;
   withSearchInput?: boolean;
   needIconCheck?: boolean;
+  pinnedOption?: SelectOption<Value>;
   callToActionButton?: CallToActionButton;
   dropdownOffset?: DropdownOffset;
   hasRightElement?: boolean;
@@ -88,6 +89,7 @@ export const Select = <Value extends SelectValue>({
   value,
   withSearchInput,
   needIconCheck,
+  pinnedOption,
   callToActionButton,
   dropdownOffset,
   hasRightElement,
@@ -97,6 +99,10 @@ export const Select = <Value extends SelectValue>({
   const [searchInputValue, setSearchInputValue] = useState('');
 
   const selectedOption = useMemo(() => {
+    if (isDefined(pinnedOption) && pinnedOption.value === value) {
+      return pinnedOption;
+    }
+
     const fromMatchingOption = options.find(
       ({ value: optionValue }) => optionValue === value,
     );
@@ -114,7 +120,7 @@ export const Select = <Value extends SelectValue>({
     }
 
     return null;
-  }, [emptyOption, options, value]);
+  }, [emptyOption, options, pinnedOption, value]);
 
   const filteredOptions = useMemo(
     () =>
@@ -129,6 +135,7 @@ export const Select = <Value extends SelectValue>({
   const isDisabled =
     disabledFromProps ||
     (options.length <= 1 &&
+      !isDefined(pinnedOption) &&
       !isDefined(callToActionButton) &&
       (!isDefined(emptyOption) || selectedOption !== emptyOption));
 
@@ -200,6 +207,25 @@ export const Select = <Value extends SelectValue>({
               {withSearchInput === true && isNonEmptyArray(filteredOptions) && (
                 <DropdownMenuSeparator />
               )}
+              {isDefined(pinnedOption) && (
+                <DropdownMenuItemsContainer scrollable={false}>
+                  <MenuItemSelect
+                    LeftIcon={pinnedOption.Icon}
+                    text={pinnedOption.label}
+                    contextualText={pinnedOption.contextualText}
+                    selected={selectedOption.value === pinnedOption.value}
+                    needIconCheck={needIconCheck}
+                    onClick={() => {
+                      onChange?.(pinnedOption.value);
+                      onBlur?.();
+                      closeDropdown(dropdownId);
+                    }}
+                  />
+                </DropdownMenuItemsContainer>
+              )}
+              {isDefined(pinnedOption) && isNonEmptyArray(filteredOptions) && (
+                <DropdownMenuSeparator />
+              )}
               {isNonEmptyArray(filteredOptions) && (
                 <DropdownMenuItemsContainer hasMaxHeight>
                   <SelectableList
@@ -220,6 +246,7 @@ export const Select = <Value extends SelectValue>({
                         <MenuItemSelect
                           LeftIcon={option.Icon}
                           text={option.label}
+                          contextualText={option.contextualText}
                           selected={selectedOption.value === option.value}
                           focused={selectedItemId === option.label}
                           needIconCheck={needIconCheck}
diff --git a/packages/twenty-front/src/modules/ui/input/components/SelectControl.tsx b/packages/twenty-front/src/modules/ui/input/components/SelectControl.tsx
index 2fbab01ed6b32..992f52dd36049 100644
--- a/packages/twenty-front/src/modules/ui/input/components/SelectControl.tsx
+++ b/packages/twenty-front/src/modules/ui/input/components/SelectControl.tsx
@@ -94,7 +94,13 @@ export const SelectControl = ({
           stroke={theme.icon.stroke.sm}
         />
       ) : null}
-      <OverflowingTextWithTooltip text={selectedOption.label} />
+      <OverflowingTextWithTooltip
+        text={
+          selectedOption.contextualText
+            ? `${selectedOption.label} · ${selectedOption.contextualText}`
+            : selectedOption.label
+        }
+      />
       <StyledIconChevronDownWrapper disabled={isDisabled}>
         <IconChevronDown size={theme.icon.size.md} />
       </StyledIconChevronDownWrapper>
diff --git a/packages/twenty-front/src/pages/settings/ai/components/SettingsAIModelsTab.tsx b/packages/twenty-front/src/pages/settings/ai/components/SettingsAIModelsTab.tsx
index f79ff8e176b45..9bb03231a27f8 100644
--- a/packages/twenty-front/src/pages/settings/ai/components/SettingsAIModelsTab.tsx
+++ b/packages/twenty-front/src/pages/settings/ai/components/SettingsAIModelsTab.tsx
@@ -1,8 +1,11 @@
 import { useState } from 'react';
 import { styled } from '@linaria/react';
 
-import { DEFAULT_FAST_MODEL } from '@/ai/constants/DefaultFastModel';
-import { DEFAULT_SMART_MODEL } from '@/ai/constants/DefaultSmartModel';
+import {
+  AUTO_SELECT_FAST_MODEL_ID,
+  AUTO_SELECT_SMART_MODEL_ID,
+} from 'twenty-shared/constants';
+
 import { useWorkspaceAiModelAvailability } from '@/ai/hooks/useWorkspaceAiModelAvailability';
 import { currentWorkspaceState } from '@/auth/states/currentWorkspaceState';
 import { aiModelsState } from '@/client-config/states/aiModelsState';
@@ -51,46 +54,44 @@ export const SettingsAIModelsTab = () => {
   const currentSmartModel = currentWorkspace?.smartModel;
   const currentFastModel = currentWorkspace?.fastModel;
 
-  const buildVirtualModelOption = (virtualModelId: string) => {
-    const virtualModel = aiModels.find(
-      (model) => model.modelId === virtualModelId,
+  const buildPinnedOption = (autoSelectModelId: string) => {
+    const autoSelectEntry = aiModels.find(
+      (model) => model.modelId === autoSelectModelId,
     );
 
-    return virtualModel
-      ? {
-          value: virtualModelId,
-          label: virtualModel.label,
-          Icon: IconTwentyStar,
-        }
-      : null;
-  };
-
-  const smartAutoOption = buildVirtualModelOption(DEFAULT_SMART_MODEL);
-  const fastAutoOption = buildVirtualModelOption(DEFAULT_FAST_MODEL);
-
-  const modelOptions = enabledModels.map((model) => {
-    const residencyFlag = model.dataResidency
-      ? ` ${getDataResidencyDisplay(model.dataResidency)}`
-      : '';
+    if (!autoSelectEntry) {
+      return undefined;
+    }
 
     return {
-      value: model.modelId,
-      label: `${model.label}${residencyFlag}`,
-      Icon: getModelIcon(model.modelFamily, model.providerName),
+      value: autoSelectModelId,
+      label: autoSelectEntry.label,
+      Icon: getModelIcon(
+        autoSelectEntry.modelFamily,
+        autoSelectEntry.providerName,
+      ),
+      contextualText: t`Best`,
     };
-  });
+  };
 
-  const smartModelOptions = [...modelOptions];
+  const smartPinnedOption = buildPinnedOption(AUTO_SELECT_SMART_MODEL_ID);
+  const fastPinnedOption = buildPinnedOption(AUTO_SELECT_FAST_MODEL_ID);
 
-  if (smartAutoOption !== null) {
-    smartModelOptions.unshift(smartAutoOption);
-  }
+  const buildModelOptions = () =>
+    enabledModels.map((model) => {
+      const residencyFlag = model.dataResidency
+        ? ` ${getDataResidencyDisplay(model.dataResidency)}`
+        : '';
 
-  const fastModelOptions = [...modelOptions];
+      return {
+        value: model.modelId,
+        label: `${model.label}${residencyFlag}`,
+        Icon: getModelIcon(model.modelFamily, model.providerName),
+      };
+    });
 
-  if (fastAutoOption !== null) {
-    fastModelOptions.unshift(fastAutoOption);
-  }
+  const smartModelOptions = buildModelOptions();
+  const fastModelOptions = buildModelOptions();
 
   const handleModelFieldChange = async (
     field: 'smartModel' | 'fastModel',
@@ -241,6 +242,7 @@ export const SettingsAIModelsTab = () => {
               value={currentSmartModel}
               onChange={(value) => handleModelFieldChange('smartModel', value)}
               options={smartModelOptions}
+              pinnedOption={smartPinnedOption}
               selectSizeVariant="small"
               dropdownWidth={GenericDropdownContentWidth.ExtraLarge}
             />
@@ -255,6 +257,7 @@ export const SettingsAIModelsTab = () => {
               value={currentFastModel}
               onChange={(value) => handleModelFieldChange('fastModel', value)}
               options={fastModelOptions}
+              pinnedOption={fastPinnedOption}
               selectSizeVariant="small"
               dropdownWidth={GenericDropdownContentWidth.ExtraLarge}
             />
diff --git a/packages/twenty-front/src/pages/settings/profile/appearance/components/DateTimeSettingsCalendarStartDaySelect.tsx b/packages/twenty-front/src/pages/settings/profile/appearance/components/DateTimeSettingsCalendarStartDaySelect.tsx
index 77a3591963fcb..af874e42f820a 100644
--- a/packages/twenty-front/src/pages/settings/profile/appearance/components/DateTimeSettingsCalendarStartDaySelect.tsx
+++ b/packages/twenty-front/src/pages/settings/profile/appearance/components/DateTimeSettingsCalendarStartDaySelect.tsx
@@ -2,7 +2,6 @@ import { useMemo } from 'react';
 
 import { detectCalendarStartDay } from '@/localization/utils/detection/detectCalendarStartDay';
 import { Select } from '@/ui/input/components/Select';
-import { type DayNameWithIndex } from '@/ui/input/components/internal/date/types/DayNameWithIndex';
 import { t } from '@lingui/core/macro';
 import { CalendarStartDay } from 'twenty-shared/constants';
 import { type SelectOption } from 'twenty-ui/input';
@@ -18,29 +17,21 @@ export const DateTimeSettingsCalendarStartDaySelect = ({
 }: DateTimeSettingsCalendarStartDaySelectProps) => {
   const systemCalendarStartDay = CalendarStartDay[detectCalendarStartDay()];
 
-  const options: SelectOption<CalendarStartDay>[] = useMemo(() => {
-    const systemDayLabel =
-      systemCalendarStartDay === CalendarStartDay.SUNDAY
-        ? t`System settings - Sunday`
-        : systemCalendarStartDay === CalendarStartDay.MONDAY
-          ? t`System settings - Monday`
-          : t`System settings - Saturday`;
+  const systemDayContextualText =
+    systemCalendarStartDay === CalendarStartDay.SUNDAY
+      ? t`Sunday`
+      : systemCalendarStartDay === CalendarStartDay.MONDAY
+        ? t`Monday`
+        : t`Saturday`;
 
-    const allowedDaysWeek: DayNameWithIndex[] = [
-      {
-        day: systemDayLabel,
-        index: CalendarStartDay.SYSTEM,
-      },
-      { day: t`Sunday`, index: CalendarStartDay.SUNDAY },
-      { day: t`Monday`, index: CalendarStartDay.MONDAY },
-      { day: t`Saturday`, index: CalendarStartDay.SATURDAY },
-    ];
-
-    return allowedDaysWeek.map(({ day, index }) => ({
-      label: day,
-      value: index as CalendarStartDay,
-    }));
-  }, [systemCalendarStartDay]);
+  const options: SelectOption<CalendarStartDay>[] = useMemo(
+    () => [
+      { label: t`Sunday`, value: CalendarStartDay.SUNDAY },
+      { label: t`Monday`, value: CalendarStartDay.MONDAY },
+      { label: t`Saturday`, value: CalendarStartDay.SATURDAY },
+    ],
+    [],
+  );
 
   return (
     <Select
@@ -50,6 +41,11 @@ export const DateTimeSettingsCalendarStartDaySelect = ({
       fullWidth
       dropdownWidthAuto
       value={value}
+      pinnedOption={{
+        label: t`System settings`,
+        value: CalendarStartDay.SYSTEM,
+        contextualText: systemDayContextualText,
+      }}
       options={options}
       onChange={onChange}
     />
diff --git a/packages/twenty-front/src/testing/mock-data/users.ts b/packages/twenty-front/src/testing/mock-data/users.ts
index bd1279262db8a..2bc2f042a5667 100644
--- a/packages/twenty-front/src/testing/mock-data/users.ts
+++ b/packages/twenty-front/src/testing/mock-data/users.ts
@@ -1,5 +1,7 @@
-import { DEFAULT_FAST_MODEL } from '@/ai/constants/DefaultFastModel';
-import { DEFAULT_SMART_MODEL } from '@/ai/constants/DefaultSmartModel';
+import {
+  AUTO_SELECT_FAST_MODEL_ID,
+  AUTO_SELECT_SMART_MODEL_ID,
+} from 'twenty-shared/constants';
 import { type CurrentUserWorkspace } from '@/auth/states/currentUserWorkspaceState';
 import { CUSTOM_WORKSPACE_APPLICATION_MOCK } from '@/object-metadata/hooks/__tests__/constants/CustomWorkspaceApplicationMock.test.constant';
 import { type WorkspaceMember } from '@/workspace-member/types/WorkspaceMember';
@@ -86,8 +88,8 @@ export const mockCurrentWorkspace = {
   updatedAt: '2023-04-26T10:23:42.33625+00:00',
   metadataVersion: 1,
   trashRetentionDays: 14,
-  fastModel: DEFAULT_FAST_MODEL,
-  smartModel: DEFAULT_SMART_MODEL,
+  fastModel: AUTO_SELECT_FAST_MODEL_ID,
+  smartModel: AUTO_SELECT_SMART_MODEL_ID,
   routerModel: 'auto',
   enabledAiModelIds: [],
   useRecommendedModels: true,
diff --git a/packages/twenty-server/src/database/typeorm/core/migrations/common/1773900000000-migrate-model-ids-to-composite-format.ts b/packages/twenty-server/src/database/typeorm/core/migrations/common/1773900000000-migrate-model-ids-to-composite-format.ts
index e4ae30003719f..d6f689755cfcf 100644
--- a/packages/twenty-server/src/database/typeorm/core/migrations/common/1773900000000-migrate-model-ids-to-composite-format.ts
+++ b/packages/twenty-server/src/database/typeorm/core/migrations/common/1773900000000-migrate-model-ids-to-composite-format.ts
@@ -6,7 +6,7 @@ export class MigrateModelIdsToCompositeFormat1773900000000
   name = 'MigrateModelIdsToCompositeFormat1773900000000';
 
   public async up(queryRunner: QueryRunner): Promise<void> {
-    // Reset workspace model columns to sentinel defaults.
+    // Reset workspace model columns to auto-select placeholders.
     // The runtime resolves these dynamically from admin preferences.
     await queryRunner.query(
       `UPDATE "core"."workspace"
@@ -25,7 +25,7 @@ export class MigrateModelIdsToCompositeFormat1773900000000
           OR array_length("enabledAiModelIds", 1) > 0`,
     );
 
-    // Reset agent model IDs to the sentinel so they fall back to workspace defaults
+    // Reset agent model IDs to the auto-select placeholder so they fall back to workspace defaults
     await queryRunner.query(
       `UPDATE "core"."agent"
        SET "modelId" = 'default-smart-model'
@@ -34,6 +34,6 @@ export class MigrateModelIdsToCompositeFormat1773900000000
   }
 
   public async down(_queryRunner: QueryRunner): Promise<void> {
-    // No reversal needed — sentinel defaults are safe to leave in place.
+    // No reversal needed — auto-select placeholders are safe to leave in place.
   }
 }
diff --git a/packages/twenty-server/src/engine/core-modules/application/utils/from-agent-manifest-to-universal-flat-agent.util.ts b/packages/twenty-server/src/engine/core-modules/application/utils/from-agent-manifest-to-universal-flat-agent.util.ts
index b9900aedb2e37..1ca6978486290 100644
--- a/packages/twenty-server/src/engine/core-modules/application/utils/from-agent-manifest-to-universal-flat-agent.util.ts
+++ b/packages/twenty-server/src/engine/core-modules/application/utils/from-agent-manifest-to-universal-flat-agent.util.ts
@@ -1,6 +1,6 @@
 import { type AgentManifest } from 'twenty-shared/application';
 
-import { DEFAULT_SMART_MODEL } from 'src/engine/metadata-modules/ai/ai-models/types/default-smart-model.const';
+import { AUTO_SELECT_SMART_MODEL_ID } from 'twenty-shared/constants';
 import { type ModelId } from 'src/engine/metadata-modules/ai/ai-models/types/model-id.type';
 import { type UniversalFlatAgent } from 'src/engine/workspace-manager/workspace-migration/universal-flat-entity/types/universal-flat-agent.type';
 
@@ -21,7 +21,7 @@ export const fromAgentManifestToUniversalFlatAgent = ({
     icon: agentManifest.icon ?? null,
     description: agentManifest.description ?? null,
     prompt: agentManifest.prompt,
-    modelId: (agentManifest.modelId as ModelId) ?? DEFAULT_SMART_MODEL,
+    modelId: (agentManifest.modelId as ModelId) ?? AUTO_SELECT_SMART_MODEL_ID,
     responseFormat: { type: 'text' },
     modelConfiguration: null,
     evaluationInputs: [],
diff --git a/packages/twenty-server/src/engine/core-modules/client-config/services/client-config.service.ts b/packages/twenty-server/src/engine/core-modules/client-config/services/client-config.service.ts
index fedd819dbdc31..ff1ea346b85f0 100644
--- a/packages/twenty-server/src/engine/core-modules/client-config/services/client-config.service.ts
+++ b/packages/twenty-server/src/engine/core-modules/client-config/services/client-config.service.ts
@@ -20,8 +20,10 @@ import { DomainServerConfigService } from 'src/engine/core-modules/domain/domain
 import { PUBLIC_FEATURE_FLAGS } from 'src/engine/core-modules/feature-flag/constants/public-feature-flag.const';
 import { TwentyConfigService } from 'src/engine/core-modules/twenty-config/twenty-config.service';
 import { convertDollarsToBillingCredits } from 'src/engine/metadata-modules/ai/ai-billing/utils/convert-dollars-to-billing-credits.util';
-import { DEFAULT_FAST_MODEL } from 'src/engine/metadata-modules/ai/ai-models/types/default-fast-model.const';
-import { DEFAULT_SMART_MODEL } from 'src/engine/metadata-modules/ai/ai-models/types/default-smart-model.const';
+import {
+  AUTO_SELECT_FAST_MODEL_ID,
+  AUTO_SELECT_SMART_MODEL_ID,
+} from 'twenty-shared/constants';
 import { MODEL_FAMILY_LABELS } from 'src/engine/metadata-modules/ai/ai-models/constants/model-family-labels.const';
 import { AiModelRegistryService } from 'src/engine/metadata-modules/ai/ai-models/services/ai-model-registry.service';
 
@@ -107,10 +109,6 @@ export class ClientConfigService {
         this.aiModelRegistryService.getDefaultSpeedModel();
       const defaultSpeedModelConfig =
         this.aiModelRegistryService.getModelConfig(defaultSpeedModel?.modelId);
-      const defaultSpeedModelLabel =
-        defaultSpeedModelConfig?.label ||
-        defaultSpeedModel?.modelId ||
-        'Default';
 
       const defaultPerformanceModel =
         this.aiModelRegistryService.getDefaultPerformanceModel();
@@ -118,23 +116,29 @@ export class ClientConfigService {
         this.aiModelRegistryService.getModelConfig(
           defaultPerformanceModel?.modelId,
         );
-      const defaultPerformanceModelLabel =
-        defaultPerformanceModelConfig?.label ||
-        defaultPerformanceModel?.modelId ||
-        'Default';
 
       aiModels.unshift(
         {
-          modelId: DEFAULT_SMART_MODEL,
-          label: `Best (${defaultPerformanceModelLabel})`,
-          sdkPackage: null,
+          modelId: AUTO_SELECT_SMART_MODEL_ID,
+          label:
+            defaultPerformanceModelConfig?.label ||
+            defaultPerformanceModel?.modelId ||
+            'Default',
+          modelFamily: defaultPerformanceModelConfig?.modelFamily,
+          providerName: defaultPerformanceModel?.providerName,
+          sdkPackage: defaultPerformanceModel?.sdkPackage ?? null,
           inputCostPerMillionTokensInCredits: 0,
           outputCostPerMillionTokensInCredits: 0,
         },
         {
-          modelId: DEFAULT_FAST_MODEL,
-          label: `Best (${defaultSpeedModelLabel})`,
-          sdkPackage: null,
+          modelId: AUTO_SELECT_FAST_MODEL_ID,
+          label:
+            defaultSpeedModelConfig?.label ||
+            defaultSpeedModel?.modelId ||
+            'Default',
+          modelFamily: defaultSpeedModelConfig?.modelFamily,
+          providerName: defaultSpeedModel?.providerName,
+          sdkPackage: defaultSpeedModel?.sdkPackage ?? null,
           inputCostPerMillionTokensInCredits: 0,
           outputCostPerMillionTokensInCredits: 0,
         },
diff --git a/packages/twenty-server/src/engine/core-modules/workspace/workspace.entity.ts b/packages/twenty-server/src/engine/core-modules/workspace/workspace.entity.ts
index dcd4807f0fe4a..43edf5ed68099 100644
--- a/packages/twenty-server/src/engine/core-modules/workspace/workspace.entity.ts
+++ b/packages/twenty-server/src/engine/core-modules/workspace/workspace.entity.ts
@@ -34,8 +34,10 @@ import { PublicDomainEntity } from 'src/engine/core-modules/public-domain/public
 import { WorkspaceSSOIdentityProviderEntity } from 'src/engine/core-modules/sso/workspace-sso-identity-provider.entity';
 import { UserWorkspaceEntity } from 'src/engine/core-modules/user-workspace/user-workspace.entity';
 import { AgentEntity } from 'src/engine/metadata-modules/ai/ai-agent/entities/agent.entity';
-import { DEFAULT_FAST_MODEL } from 'src/engine/metadata-modules/ai/ai-models/types/default-fast-model.const';
-import { DEFAULT_SMART_MODEL } from 'src/engine/metadata-modules/ai/ai-models/types/default-smart-model.const';
+import {
+  AUTO_SELECT_FAST_MODEL_ID,
+  AUTO_SELECT_SMART_MODEL_ID,
+} from 'twenty-shared/constants';
 import { type ModelId } from 'src/engine/metadata-modules/ai/ai-models/types/model-id.type';
 import { RoleDTO } from 'src/engine/metadata-modules/role/dtos/role.dto';
 import { ViewFieldDTO } from 'src/engine/metadata-modules/view-field/dtos/view-field.dto';
@@ -301,11 +303,11 @@ export class WorkspaceEntity {
   version: string | null;
 
   @Field(() => String, { nullable: false })
-  @Column({ type: 'varchar', nullable: false, default: DEFAULT_FAST_MODEL })
+  @Column({ type: 'varchar', nullable: false, default: AUTO_SELECT_FAST_MODEL_ID })
   fastModel: ModelId;
 
   @Field(() => String, { nullable: false })
-  @Column({ type: 'varchar', nullable: false, default: DEFAULT_SMART_MODEL })
+  @Column({ type: 'varchar', nullable: false, default: AUTO_SELECT_SMART_MODEL_ID })
   smartModel: ModelId;
 
   @Field(() => String, { nullable: true })
diff --git a/packages/twenty-server/src/engine/metadata-modules/ai/ai-agent/entities/agent.entity.ts b/packages/twenty-server/src/engine/metadata-modules/ai/ai-agent/entities/agent.entity.ts
index 43336d6e3d327..9145b02874154 100644
--- a/packages/twenty-server/src/engine/metadata-modules/ai/ai-agent/entities/agent.entity.ts
+++ b/packages/twenty-server/src/engine/metadata-modules/ai/ai-agent/entities/agent.entity.ts
@@ -10,7 +10,7 @@ import {
 
 import { AgentResponseFormat } from 'src/engine/metadata-modules/ai/ai-agent/types/agent-response-format.type';
 import { ModelConfiguration } from 'src/engine/metadata-modules/ai/ai-agent/types/modelConfiguration';
-import { DEFAULT_SMART_MODEL } from 'src/engine/metadata-modules/ai/ai-models/types/default-smart-model.const';
+import { AUTO_SELECT_SMART_MODEL_ID } from 'twenty-shared/constants';
 import { type ModelId } from 'src/engine/metadata-modules/ai/ai-models/types/model-id.type';
 import { SyncableEntity } from 'src/engine/workspace-manager/types/syncable-entity.interface';
 import { JsonbProperty } from 'src/engine/workspace-manager/workspace-migration/universal-flat-entity/types/jsonb-property.type';
@@ -43,7 +43,7 @@ export class AgentEntity
   @Column({ nullable: false, type: 'text' })
   prompt: string;
 
-  @Column({ nullable: false, type: 'varchar', default: DEFAULT_SMART_MODEL })
+  @Column({ nullable: false, type: 'varchar', default: AUTO_SELECT_SMART_MODEL_ID })
   modelId: ModelId;
 
   // Should not be nullable
diff --git a/packages/twenty-server/src/engine/metadata-modules/ai/ai-chat/controllers/agent-chat.controller.ts b/packages/twenty-server/src/engine/metadata-modules/ai/ai-chat/controllers/agent-chat.controller.ts
index 8dfb98ccc235f..e80a6ff49c830 100644
--- a/packages/twenty-server/src/engine/metadata-modules/ai/ai-chat/controllers/agent-chat.controller.ts
+++ b/packages/twenty-server/src/engine/metadata-modules/ai/ai-chat/controllers/agent-chat.controller.ts
@@ -39,9 +39,9 @@ import { AiModelRegistryService } from 'src/engine/metadata-modules/ai/ai-models
 @Controller('rest/agent-chat')
 @UseGuards(JwtAuthGuard, WorkspaceAuthGuard)
 @UseFilters(
+  RestApiExceptionFilter,
   AgentRestApiExceptionFilter,
   BillingRestApiExceptionFilter,
-  RestApiExceptionFilter,
 )
 export class AgentChatController {
   constructor(
@@ -59,6 +59,7 @@ export class AgentChatController {
       threadId: string;
       messages: ExtendedUIMessage[];
       browsingContext?: BrowsingContextType | null;
+      modelId?: string;
     },
     @AuthUserWorkspaceId() userWorkspaceId: string,
     @AuthWorkspace() workspace: WorkspaceEntity,
@@ -71,7 +72,7 @@ export class AgentChatController {
       );
     }
 
-    const resolvedModelId = workspace.smartModel;
+    const resolvedModelId = body.modelId ?? workspace.smartModel;
 
     this.aiModelRegistryService.validateModelAvailability(
       resolvedModelId,
@@ -96,6 +97,7 @@ export class AgentChatController {
       threadId: body.threadId,
       messages: body.messages,
       browsingContext: body.browsingContext ?? null,
+      modelId: body.modelId,
       userWorkspaceId,
       workspace,
       response,
diff --git a/packages/twenty-server/src/engine/metadata-modules/ai/ai-chat/services/agent-chat-streaming.service.ts b/packages/twenty-server/src/engine/metadata-modules/ai/ai-chat/services/agent-chat-streaming.service.ts
index 34f2647319f8e..ffb6905406f2d 100644
--- a/packages/twenty-server/src/engine/metadata-modules/ai/ai-chat/services/agent-chat-streaming.service.ts
+++ b/packages/twenty-server/src/engine/metadata-modules/ai/ai-chat/services/agent-chat-streaming.service.ts
@@ -33,6 +33,7 @@ export type StreamAgentChatOptions = {
   response: Response;
   messages: ExtendedUIMessage[];
   browsingContext: BrowsingContextType | null;
+  modelId?: string;
 };
 
 @Injectable()
@@ -51,6 +52,7 @@ export class AgentChatStreamingService {
     messages,
     browsingContext,
     response,
+    modelId,
   }: StreamAgentChatOptions) {
     const thread = await this.threadRepository.findOne({
       where: {
@@ -114,6 +116,7 @@ export class AgentChatStreamingService {
               messages,
               browsingContext,
               onCodeExecutionUpdate,
+              modelId,
             });
 
           let streamUsage = {
diff --git a/packages/twenty-server/src/engine/metadata-modules/ai/ai-chat/services/chat-execution.service.ts b/packages/twenty-server/src/engine/metadata-modules/ai/ai-chat/services/chat-execution.service.ts
index 18615417e583e..bbcdfce08824a 100644
--- a/packages/twenty-server/src/engine/metadata-modules/ai/ai-chat/services/chat-execution.service.ts
+++ b/packages/twenty-server/src/engine/metadata-modules/ai/ai-chat/services/chat-execution.service.ts
@@ -60,6 +60,7 @@ export type ChatExecutionOptions = {
   messages: UIMessage<unknown, UIDataTypes, UITools>[];
   browsingContext: BrowsingContextType | null;
   onCodeExecutionUpdate?: CodeExecutionStreamEmitter;
+  modelId?: string;
 };
 
 export type ChatExecutionResult = {
@@ -89,6 +90,7 @@ export class ChatExecutionService {
     messages,
     browsingContext,
     onCodeExecutionUpdate,
+    modelId,
   }: ChatExecutionOptions): Promise<ChatExecutionResult> {
     const { actorContext, roleId, userId, userContext } =
       await this.agentActorContextService.buildUserAndAgentActorContext(
@@ -128,13 +130,16 @@ export class ChatExecutionService {
       toolContext,
     );
 
-    const modelId = workspace.smartModel;
+    const resolvedModelId = modelId ?? workspace.smartModel;
 
-    this.aiModelRegistryService.validateModelAvailability(modelId, workspace);
+    this.aiModelRegistryService.validateModelAvailability(
+      resolvedModelId,
+      workspace,
+    );
 
     const registeredModel =
       await this.aiModelRegistryService.resolveModelForAgent({
-        modelId,
+        modelId: resolvedModelId,
       });
 
     const modelConfig = this.aiModelRegistryService.getEffectiveModelConfig(
diff --git a/packages/twenty-server/src/engine/metadata-modules/ai/ai-models/ai-providers.json b/packages/twenty-server/src/engine/metadata-modules/ai/ai-models/ai-providers.json
index 4aaf0d291b377..b42010f22ea0a 100644
--- a/packages/twenty-server/src/engine/metadata-modules/ai/ai-models/ai-providers.json
+++ b/packages/twenty-server/src/engine/metadata-modules/ai/ai-models/ai-providers.json
@@ -7,7 +7,7 @@
       {
         "name": "gpt-4o-2024-11-20",
         "label": "GPT-4o (2024-11-20)",
-        "modelFamily": "gpt",
+        "modelFamily": "GPT",
         "inputCostPerMillionTokens": 2.5,
         "outputCostPerMillionTokens": 10,
         "cachedInputCostPerMillionTokens": 1.25,
@@ -18,7 +18,7 @@
       {
         "name": "gpt-5.3-codex",
         "label": "GPT-5.3 Codex",
-        "modelFamily": "gpt",
+        "modelFamily": "GPT",
         "inputCostPerMillionTokens": 1.75,
         "outputCostPerMillionTokens": 14,
         "cachedInputCostPerMillionTokens": 0.175,
@@ -30,7 +30,7 @@
       {
         "name": "gpt-5-codex",
         "label": "GPT-5-Codex",
-        "modelFamily": "gpt",
+        "modelFamily": "GPT",
         "inputCostPerMillionTokens": 1.25,
         "outputCostPerMillionTokens": 10,
         "cachedInputCostPerMillionTokens": 0.125,
@@ -42,7 +42,7 @@
       {
         "name": "gpt-5-pro",
         "label": "GPT-5 Pro",
-        "modelFamily": "gpt",
+        "modelFamily": "GPT",
         "inputCostPerMillionTokens": 15,
         "outputCostPerMillionTokens": 120,
         "contextWindowTokens": 400000,
@@ -53,7 +53,7 @@
       {
         "name": "gpt-4o-mini",
         "label": "GPT-4o mini",
-        "modelFamily": "gpt",
+        "modelFamily": "GPT",
         "inputCostPerMillionTokens": 0.15,
         "outputCostPerMillionTokens": 0.6,
         "cachedInputCostPerMillionTokens": 0.08,
@@ -64,7 +64,7 @@
       {
         "name": "codex-mini-latest",
         "label": "Codex Mini",
-        "modelFamily": "gpt",
+        "modelFamily": "GPT",
         "inputCostPerMillionTokens": 1.5,
         "outputCostPerMillionTokens": 6,
         "cachedInputCostPerMillionTokens": 0.375,
@@ -75,7 +75,7 @@
       {
         "name": "gpt-5.1-codex-max",
         "label": "GPT-5.1 Codex Max",
-        "modelFamily": "gpt",
+        "modelFamily": "GPT",
         "inputCostPerMillionTokens": 1.25,
         "outputCostPerMillionTokens": 10,
         "cachedInputCostPerMillionTokens": 0.125,
@@ -87,7 +87,7 @@
       {
         "name": "gpt-4o-2024-05-13",
         "label": "GPT-4o (2024-05-13)",
-        "modelFamily": "gpt",
+        "modelFamily": "GPT",
         "inputCostPerMillionTokens": 5,
         "outputCostPerMillionTokens": 15,
         "contextWindowTokens": 128000,
@@ -97,7 +97,7 @@
       {
         "name": "gpt-5.2-chat-latest",
         "label": "GPT-5.2 Chat",
-        "modelFamily": "gpt",
+        "modelFamily": "GPT",
         "inputCostPerMillionTokens": 1.75,
         "outputCostPerMillionTokens": 14,
         "cachedInputCostPerMillionTokens": 0.175,
@@ -109,7 +109,7 @@
       {
         "name": "gpt-5.2-codex",
         "label": "GPT-5.2 Codex",
-        "modelFamily": "gpt",
+        "modelFamily": "GPT",
         "inputCostPerMillionTokens": 1.75,
         "outputCostPerMillionTokens": 14,
         "cachedInputCostPerMillionTokens": 0.175,
@@ -121,7 +121,7 @@
       {
         "name": "o3-deep-research",
         "label": "o3-deep-research",
-        "modelFamily": "gpt",
+        "modelFamily": "GPT",
         "inputCostPerMillionTokens": 10,
         "outputCostPerMillionTokens": 40,
         "cachedInputCostPerMillionTokens": 2.5,
@@ -133,7 +133,7 @@
       {
         "name": "o1",
         "label": "o1",
-        "modelFamily": "gpt",
+        "modelFamily": "GPT",
         "inputCostPerMillionTokens": 15,
         "outputCostPerMillionTokens": 60,
         "cachedInputCostPerMillionTokens": 7.5,
@@ -145,7 +145,7 @@
       {
         "name": "gpt-5.1",
         "label": "GPT-5.1",
-        "modelFamily": "gpt",
+        "modelFamily": "GPT",
         "inputCostPerMillionTokens": 1.25,
         "outputCostPerMillionTokens": 10,
         "cachedInputCostPerMillionTokens": 0.13,
@@ -157,7 +157,7 @@
       {
         "name": "o4-mini-deep-research",
         "label": "o4-mini-deep-research",
-        "modelFamily": "gpt",
+        "modelFamily": "GPT",
         "inputCostPerMillionTokens": 2,
         "outputCostPerMillionTokens": 8,
         "cachedInputCostPerMillionTokens": 0.5,
@@ -169,7 +169,7 @@
       {
         "name": "gpt-5.3-codex-spark",
         "label": "GPT-5.3 Codex Spark",
-        "modelFamily": "gpt",
+        "modelFamily": "GPT",
         "inputCostPerMillionTokens": 1.75,
         "outputCostPerMillionTokens": 14,
         "cachedInputCostPerMillionTokens": 0.175,
@@ -181,7 +181,7 @@
       {
         "name": "o3",
         "label": "o3",
-        "modelFamily": "gpt",
+        "modelFamily": "GPT",
         "inputCostPerMillionTokens": 2,
         "outputCostPerMillionTokens": 8,
         "cachedInputCostPerMillionTokens": 0.5,
@@ -193,7 +193,7 @@
       {
         "name": "gpt-4.1-nano",
         "label": "GPT-4.1 nano",
-        "modelFamily": "gpt",
+        "modelFamily": "GPT",
         "inputCostPerMillionTokens": 0.1,
         "outputCostPerMillionTokens": 0.4,
         "cachedInputCostPerMillionTokens": 0.03,
@@ -204,7 +204,7 @@
       {
         "name": "gpt-5.1-codex-mini",
         "label": "GPT-5.1 Codex mini",
-        "modelFamily": "gpt",
+        "modelFamily": "GPT",
         "inputCostPerMillionTokens": 0.25,
         "outputCostPerMillionTokens": 2,
         "cachedInputCostPerMillionTokens": 0.025,
@@ -216,7 +216,7 @@
       {
         "name": "gpt-5.2",
         "label": "GPT-5.2",
-        "modelFamily": "gpt",
+        "modelFamily": "GPT",
         "inputCostPerMillionTokens": 1.75,
         "outputCostPerMillionTokens": 14,
         "cachedInputCostPerMillionTokens": 0.175,
@@ -228,7 +228,7 @@
       {
         "name": "gpt-4.1",
         "label": "GPT-4.1",
-        "modelFamily": "gpt",
+        "modelFamily": "GPT",
         "inputCostPerMillionTokens": 2,
         "outputCostPerMillionTokens": 8,
         "cachedInputCostPerMillionTokens": 0.5,
@@ -239,7 +239,7 @@
       {
         "name": "o3-pro",
         "label": "o3-pro",
-        "modelFamily": "gpt",
+        "modelFamily": "GPT",
         "inputCostPerMillionTokens": 20,
         "outputCostPerMillionTokens": 80,
         "contextWindowTokens": 200000,
@@ -250,7 +250,7 @@
       {
         "name": "gpt-4-turbo",
         "label": "GPT-4 Turbo",
-        "modelFamily": "gpt",
+        "modelFamily": "GPT",
         "inputCostPerMillionTokens": 10,
         "outputCostPerMillionTokens": 30,
         "contextWindowTokens": 128000,
@@ -260,7 +260,7 @@
       {
         "name": "gpt-5",
         "label": "GPT-5",
-        "modelFamily": "gpt",
+        "modelFamily": "GPT",
         "inputCostPerMillionTokens": 1.25,
         "outputCostPerMillionTokens": 10,
         "cachedInputCostPerMillionTokens": 0.125,
@@ -272,7 +272,7 @@
       {
         "name": "o4-mini",
         "label": "o4-mini",
-        "modelFamily": "gpt",
+        "modelFamily": "GPT",
         "inputCostPerMillionTokens": 1.1,
         "outputCostPerMillionTokens": 4.4,
         "cachedInputCostPerMillionTokens": 0.28,
@@ -284,7 +284,7 @@
       {
         "name": "gpt-4.1-mini",
         "label": "GPT-4.1 mini",
-        "modelFamily": "gpt",
+        "modelFamily": "GPT",
         "inputCostPerMillionTokens": 0.4,
         "outputCostPerMillionTokens": 1.6,
         "cachedInputCostPerMillionTokens": 0.1,
@@ -295,7 +295,7 @@
       {
         "name": "gpt-5.4",
         "label": "GPT-5.4",
-        "modelFamily": "gpt",
+        "modelFamily": "GPT",
         "inputCostPerMillionTokens": 2.5,
         "outputCostPerMillionTokens": 15,
         "cachedInputCostPerMillionTokens": 0.25,
@@ -313,7 +313,7 @@
       {
         "name": "gpt-5.4-pro",
         "label": "GPT-5.4 Pro",
-        "modelFamily": "gpt",
+        "modelFamily": "GPT",
         "inputCostPerMillionTokens": 30,
         "outputCostPerMillionTokens": 180,
         "longContextCost": {
@@ -329,7 +329,7 @@
       {
         "name": "o1-pro",
         "label": "o1-pro",
-        "modelFamily": "gpt",
+        "modelFamily": "GPT",
         "inputCostPerMillionTokens": 150,
         "outputCostPerMillionTokens": 600,
         "contextWindowTokens": 200000,
@@ -340,7 +340,7 @@
       {
         "name": "gpt-5.1-codex",
         "label": "GPT-5.1 Codex",
-        "modelFamily": "gpt",
+        "modelFamily": "GPT",
         "inputCostPerMillionTokens": 1.25,
         "outputCostPerMillionTokens": 10,
         "cachedInputCostPerMillionTokens": 0.125,
@@ -352,7 +352,7 @@
       {
         "name": "gpt-5.2-pro",
         "label": "GPT-5.2 Pro",
-        "modelFamily": "gpt",
+        "modelFamily": "GPT",
         "inputCostPerMillionTokens": 21,
         "outputCostPerMillionTokens": 168,
         "contextWindowTokens": 400000,
@@ -363,7 +363,7 @@
       {
         "name": "o3-mini",
         "label": "o3-mini",
-        "modelFamily": "gpt",
+        "modelFamily": "GPT",
         "inputCostPerMillionTokens": 1.1,
         "outputCostPerMillionTokens": 4.4,
         "cachedInputCostPerMillionTokens": 0.55,
@@ -374,7 +374,7 @@
       {
         "name": "gpt-4o-2024-08-06",
         "label": "GPT-4o (2024-08-06)",
-        "modelFamily": "gpt",
+        "modelFamily": "GPT",
         "inputCostPerMillionTokens": 2.5,
         "outputCostPerMillionTokens": 10,
         "cachedInputCostPerMillionTokens": 1.25,
@@ -385,7 +385,7 @@
       {
         "name": "gpt-5-mini",
         "label": "GPT-5 Mini",
-        "modelFamily": "gpt",
+        "modelFamily": "GPT",
         "inputCostPerMillionTokens": 0.25,
         "outputCostPerMillionTokens": 2,
         "cachedInputCostPerMillionTokens": 0.025,
@@ -397,7 +397,7 @@
       {
         "name": "gpt-5.4-nano",
         "label": "GPT-5.4 nano",
-        "modelFamily": "gpt",
+        "modelFamily": "GPT",
         "inputCostPerMillionTokens": 0.2,
         "outputCostPerMillionTokens": 1.25,
         "cachedInputCostPerMillionTokens": 0.02,
@@ -409,7 +409,7 @@
       {
         "name": "gpt-5.1-chat-latest",
         "label": "GPT-5.1 Chat",
-        "modelFamily": "gpt",
+        "modelFamily": "GPT",
         "inputCostPerMillionTokens": 1.25,
         "outputCostPerMillionTokens": 10,
         "cachedInputCostPerMillionTokens": 0.125,
@@ -421,7 +421,7 @@
       {
         "name": "gpt-4",
         "label": "GPT-4",
-        "modelFamily": "gpt",
+        "modelFamily": "GPT",
         "inputCostPerMillionTokens": 30,
         "outputCostPerMillionTokens": 60,
         "contextWindowTokens": 8192,
@@ -430,7 +430,7 @@
       {
         "name": "gpt-5-nano",
         "label": "GPT-5 Nano",
-        "modelFamily": "gpt",
+        "modelFamily": "GPT",
         "inputCostPerMillionTokens": 0.05,
         "outputCostPerMillionTokens": 0.4,
         "cachedInputCostPerMillionTokens": 0.005,
@@ -442,7 +442,7 @@
       {
         "name": "gpt-4o",
         "label": "GPT-4o",
-        "modelFamily": "gpt",
+        "modelFamily": "GPT",
         "inputCostPerMillionTokens": 2.5,
         "outputCostPerMillionTokens": 10,
         "cachedInputCostPerMillionTokens": 1.25,
@@ -453,7 +453,7 @@
       {
         "name": "gpt-5.4-mini",
         "label": "GPT-5.4 mini",
-        "modelFamily": "gpt",
+        "modelFamily": "GPT",
         "inputCostPerMillionTokens": 0.75,
         "outputCostPerMillionTokens": 4.5,
         "cachedInputCostPerMillionTokens": 0.075,
@@ -472,7 +472,7 @@
       {
         "name": "claude-opus-4-5-20251101",
         "label": "Claude Opus 4.5",
-        "modelFamily": "claude",
+        "modelFamily": "CLAUDE",
         "inputCostPerMillionTokens": 5,
         "outputCostPerMillionTokens": 25,
         "cachedInputCostPerMillionTokens": 0.5,
@@ -485,7 +485,7 @@
       {
         "name": "claude-3-5-haiku-latest",
         "label": "Claude Haiku 3.5 (latest)",
-        "modelFamily": "claude",
+        "modelFamily": "CLAUDE",
         "inputCostPerMillionTokens": 0.8,
         "outputCostPerMillionTokens": 4,
         "cachedInputCostPerMillionTokens": 0.08,
@@ -497,7 +497,7 @@
       {
         "name": "claude-opus-4-1",
         "label": "Claude Opus 4.1 (latest)",
-        "modelFamily": "claude",
+        "modelFamily": "CLAUDE",
         "inputCostPerMillionTokens": 15,
         "outputCostPerMillionTokens": 75,
         "cachedInputCostPerMillionTokens": 1.5,
@@ -510,7 +510,7 @@
       {
         "name": "claude-3-5-sonnet-20241022",
         "label": "Claude Sonnet 3.5 v2",
-        "modelFamily": "claude",
+        "modelFamily": "CLAUDE",
         "inputCostPerMillionTokens": 3,
         "outputCostPerMillionTokens": 15,
         "cachedInputCostPerMillionTokens": 0.3,
@@ -522,7 +522,7 @@
       {
         "name": "claude-3-sonnet-20240229",
         "label": "Claude Sonnet 3",
-        "modelFamily": "claude",
+        "modelFamily": "CLAUDE",
         "inputCostPerMillionTokens": 3,
         "outputCostPerMillionTokens": 15,
         "cachedInputCostPerMillionTokens": 0.3,
@@ -534,7 +534,7 @@
       {
         "name": "claude-opus-4-6",
         "label": "Claude Opus 4.6",
-        "modelFamily": "claude",
+        "modelFamily": "CLAUDE",
         "inputCostPerMillionTokens": 5,
         "outputCostPerMillionTokens": 25,
         "cachedInputCostPerMillionTokens": 0.5,
@@ -547,7 +547,7 @@
       {
         "name": "claude-sonnet-4-6",
         "label": "Claude Sonnet 4.6",
-        "modelFamily": "claude",
+        "modelFamily": "CLAUDE",
         "inputCostPerMillionTokens": 3,
         "outputCostPerMillionTokens": 15,
         "cachedInputCostPerMillionTokens": 0.3,
@@ -560,7 +560,7 @@
       {
         "name": "claude-sonnet-4-0",
         "label": "Claude Sonnet 4 (latest)",
-        "modelFamily": "claude",
+        "modelFamily": "CLAUDE",
         "inputCostPerMillionTokens": 3,
         "outputCostPerMillionTokens": 15,
         "cachedInputCostPerMillionTokens": 0.3,
@@ -573,7 +573,7 @@
       {
         "name": "claude-opus-4-20250514",
         "label": "Claude Opus 4",
-        "modelFamily": "claude",
+        "modelFamily": "CLAUDE",
         "inputCostPerMillionTokens": 15,
         "outputCostPerMillionTokens": 75,
         "cachedInputCostPerMillionTokens": 1.5,
@@ -586,7 +586,7 @@
       {
         "name": "claude-sonnet-4-5-20250929",
         "label": "Claude Sonnet 4.5",
-        "modelFamily": "claude",
+        "modelFamily": "CLAUDE",
         "inputCostPerMillionTokens": 3,
         "outputCostPerMillionTokens": 15,
         "cachedInputCostPerMillionTokens": 0.3,
@@ -599,7 +599,7 @@
       {
         "name": "claude-opus-4-0",
         "label": "Claude Opus 4 (latest)",
-        "modelFamily": "claude",
+        "modelFamily": "CLAUDE",
         "inputCostPerMillionTokens": 15,
         "outputCostPerMillionTokens": 75,
         "cachedInputCostPerMillionTokens": 1.5,
@@ -612,7 +612,7 @@
       {
         "name": "claude-3-5-haiku-20241022",
         "label": "Claude Haiku 3.5",
-        "modelFamily": "claude",
+        "modelFamily": "CLAUDE",
         "inputCostPerMillionTokens": 0.8,
         "outputCostPerMillionTokens": 4,
         "cachedInputCostPerMillionTokens": 0.08,
@@ -624,7 +624,7 @@
       {
         "name": "claude-3-5-sonnet-20240620",
         "label": "Claude Sonnet 3.5",
-        "modelFamily": "claude",
+        "modelFamily": "CLAUDE",
         "inputCostPerMillionTokens": 3,
         "outputCostPerMillionTokens": 15,
         "cachedInputCostPerMillionTokens": 0.3,
@@ -636,7 +636,7 @@
       {
         "name": "claude-3-7-sonnet-latest",
         "label": "Claude Sonnet 3.7 (latest)",
-        "modelFamily": "claude",
+        "modelFamily": "CLAUDE",
         "inputCostPerMillionTokens": 3,
         "outputCostPerMillionTokens": 15,
         "cachedInputCostPerMillionTokens": 0.3,
@@ -649,7 +649,7 @@
       {
         "name": "claude-3-7-sonnet-20250219",
         "label": "Claude Sonnet 3.7",
-        "modelFamily": "claude",
+        "modelFamily": "CLAUDE",
         "inputCostPerMillionTokens": 3,
         "outputCostPerMillionTokens": 15,
         "cachedInputCostPerMillionTokens": 0.3,
@@ -662,7 +662,7 @@
       {
         "name": "claude-3-haiku-20240307",
         "label": "Claude Haiku 3",
-        "modelFamily": "claude",
+        "modelFamily": "CLAUDE",
         "inputCostPerMillionTokens": 0.25,
         "outputCostPerMillionTokens": 1.25,
         "cachedInputCostPerMillionTokens": 0.03,
@@ -674,7 +674,7 @@
       {
         "name": "claude-haiku-4-5-20251001",
         "label": "Claude Haiku 4.5",
-        "modelFamily": "claude",
+        "modelFamily": "CLAUDE",
         "inputCostPerMillionTokens": 1,
         "outputCostPerMillionTokens": 5,
         "cachedInputCostPerMillionTokens": 0.1,
@@ -687,7 +687,7 @@
       {
         "name": "claude-haiku-4-5",
         "label": "Claude Haiku 4.5 (latest)",
-        "modelFamily": "claude",
+        "modelFamily": "CLAUDE",
         "inputCostPerMillionTokens": 1,
         "outputCostPerMillionTokens": 5,
         "cachedInputCostPerMillionTokens": 0.1,
@@ -700,7 +700,7 @@
       {
         "name": "claude-opus-4-5",
         "label": "Claude Opus 4.5 (latest)",
-        "modelFamily": "claude",
+        "modelFamily": "CLAUDE",
         "inputCostPerMillionTokens": 5,
         "outputCostPerMillionTokens": 25,
         "cachedInputCostPerMillionTokens": 0.5,
@@ -713,7 +713,7 @@
       {
         "name": "claude-3-opus-20240229",
         "label": "Claude Opus 3",
-        "modelFamily": "claude",
+        "modelFamily": "CLAUDE",
         "inputCostPerMillionTokens": 15,
         "outputCostPerMillionTokens": 75,
         "cachedInputCostPerMillionTokens": 1.5,
@@ -725,7 +725,7 @@
       {
         "name": "claude-sonnet-4-5",
         "label": "Claude Sonnet 4.5 (latest)",
-        "modelFamily": "claude",
+        "modelFamily": "CLAUDE",
         "inputCostPerMillionTokens": 3,
         "outputCostPerMillionTokens": 15,
         "cachedInputCostPerMillionTokens": 0.3,
@@ -738,7 +738,7 @@
       {
         "name": "claude-sonnet-4-20250514",
         "label": "Claude Sonnet 4",
-        "modelFamily": "claude",
+        "modelFamily": "CLAUDE",
         "inputCostPerMillionTokens": 3,
         "outputCostPerMillionTokens": 15,
         "cachedInputCostPerMillionTokens": 0.3,
@@ -751,7 +751,7 @@
       {
         "name": "claude-opus-4-1-20250805",
         "label": "Claude Opus 4.1",
-        "modelFamily": "claude",
+        "modelFamily": "CLAUDE",
         "inputCostPerMillionTokens": 15,
         "outputCostPerMillionTokens": 75,
         "cachedInputCostPerMillionTokens": 1.5,
@@ -771,7 +771,7 @@
       {
         "name": "gemini-2.5-flash-lite-preview-09-2025",
         "label": "Gemini 2.5 Flash Lite Preview 09-25",
-        "modelFamily": "gemini",
+        "modelFamily": "GEMINI",
         "inputCostPerMillionTokens": 0.1,
         "outputCostPerMillionTokens": 0.4,
         "cachedInputCostPerMillionTokens": 0.025,
@@ -783,7 +783,7 @@
       {
         "name": "gemini-3.1-pro-preview-customtools",
         "label": "Gemini 3.1 Pro Preview Custom Tools",
-        "modelFamily": "gemini",
+        "modelFamily": "GEMINI",
         "inputCostPerMillionTokens": 2,
         "outputCostPerMillionTokens": 12,
         "cachedInputCostPerMillionTokens": 0.2,
@@ -801,7 +801,7 @@
       {
         "name": "gemini-2.5-pro-preview-06-05",
         "label": "Gemini 2.5 Pro Preview 06-05",
-        "modelFamily": "gemini",
+        "modelFamily": "GEMINI",
         "inputCostPerMillionTokens": 1.25,
         "outputCostPerMillionTokens": 10,
         "cachedInputCostPerMillionTokens": 0.31,
@@ -813,7 +813,7 @@
       {
         "name": "gemini-2.5-flash-preview-04-17",
         "label": "Gemini 2.5 Flash Preview 04-17",
-        "modelFamily": "gemini",
+        "modelFamily": "GEMINI",
         "inputCostPerMillionTokens": 0.15,
         "outputCostPerMillionTokens": 0.6,
         "cachedInputCostPerMillionTokens": 0.0375,
@@ -825,7 +825,7 @@
       {
         "name": "gemini-2.5-flash-preview-09-2025",
         "label": "Gemini 2.5 Flash Preview 09-25",
-        "modelFamily": "gemini",
+        "modelFamily": "GEMINI",
         "inputCostPerMillionTokens": 0.3,
         "outputCostPerMillionTokens": 2.5,
         "cachedInputCostPerMillionTokens": 0.075,
@@ -837,7 +837,7 @@
       {
         "name": "gemini-2.5-pro-preview-05-06",
         "label": "Gemini 2.5 Pro Preview 05-06",
-        "modelFamily": "gemini",
+        "modelFamily": "GEMINI",
         "inputCostPerMillionTokens": 1.25,
         "outputCostPerMillionTokens": 10,
         "cachedInputCostPerMillionTokens": 0.31,
@@ -849,7 +849,7 @@
       {
         "name": "gemini-2.5-flash-preview-05-20",
         "label": "Gemini 2.5 Flash Preview 05-20",
-        "modelFamily": "gemini",
+        "modelFamily": "GEMINI",
         "inputCostPerMillionTokens": 0.15,
         "outputCostPerMillionTokens": 0.6,
         "cachedInputCostPerMillionTokens": 0.0375,
@@ -861,7 +861,7 @@
       {
         "name": "gemini-2.5-flash",
         "label": "Gemini 2.5 Flash",
-        "modelFamily": "gemini",
+        "modelFamily": "GEMINI",
         "inputCostPerMillionTokens": 0.3,
         "outputCostPerMillionTokens": 2.5,
         "cachedInputCostPerMillionTokens": 0.075,
@@ -873,7 +873,7 @@
       {
         "name": "gemini-3.1-flash-lite-preview",
         "label": "Gemini 3.1 Flash Lite Preview",
-        "modelFamily": "gemini",
+        "modelFamily": "GEMINI",
         "inputCostPerMillionTokens": 0.25,
         "outputCostPerMillionTokens": 1.5,
         "cachedInputCostPerMillionTokens": 0.025,
@@ -886,7 +886,7 @@
       {
         "name": "gemini-live-2.5-flash",
         "label": "Gemini Live 2.5 Flash",
-        "modelFamily": "gemini",
+        "modelFamily": "GEMINI",
         "inputCostPerMillionTokens": 0.5,
         "outputCostPerMillionTokens": 2,
         "contextWindowTokens": 128000,
@@ -897,7 +897,7 @@
       {
         "name": "gemini-3-flash-preview",
         "label": "Gemini 3 Flash Preview",
-        "modelFamily": "gemini",
+        "modelFamily": "GEMINI",
         "inputCostPerMillionTokens": 0.5,
         "outputCostPerMillionTokens": 3,
         "cachedInputCostPerMillionTokens": 0.05,
@@ -915,7 +915,7 @@
       {
         "name": "gemini-live-2.5-flash-preview-native-audio",
         "label": "Gemini Live 2.5 Flash Preview Native Audio",
-        "modelFamily": "gemini",
+        "modelFamily": "GEMINI",
         "inputCostPerMillionTokens": 0.5,
         "outputCostPerMillionTokens": 2,
         "contextWindowTokens": 131072,
@@ -926,7 +926,7 @@
       {
         "name": "gemini-2.5-flash-lite",
         "label": "Gemini 2.5 Flash Lite",
-        "modelFamily": "gemini",
+        "modelFamily": "GEMINI",
         "inputCostPerMillionTokens": 0.1,
         "outputCostPerMillionTokens": 0.4,
         "cachedInputCostPerMillionTokens": 0.025,
@@ -938,7 +938,7 @@
       {
         "name": "gemini-3.1-pro-preview",
         "label": "Gemini 3.1 Pro Preview",
-        "modelFamily": "gemini",
+        "modelFamily": "GEMINI",
         "inputCostPerMillionTokens": 2,
         "outputCostPerMillionTokens": 12,
         "cachedInputCostPerMillionTokens": 0.2,
@@ -956,7 +956,7 @@
       {
         "name": "gemini-flash-latest",
         "label": "Gemini Flash Latest",
-        "modelFamily": "gemini",
+        "modelFamily": "GEMINI",
         "inputCostPerMillionTokens": 0.3,
         "outputCostPerMillionTokens": 2.5,
         "cachedInputCostPerMillionTokens": 0.075,
@@ -968,7 +968,7 @@
       {
         "name": "gemini-2.5-flash-lite-preview-06-17",
         "label": "Gemini 2.5 Flash Lite Preview 06-17",
-        "modelFamily": "gemini",
+        "modelFamily": "GEMINI",
         "inputCostPerMillionTokens": 0.1,
         "outputCostPerMillionTokens": 0.4,
         "cachedInputCostPerMillionTokens": 0.025,
@@ -980,7 +980,7 @@
       {
         "name": "gemini-1.5-flash-8b",
         "label": "Gemini 1.5 Flash-8B",
-        "modelFamily": "gemini",
+        "modelFamily": "GEMINI",
         "inputCostPerMillionTokens": 0.0375,
         "outputCostPerMillionTokens": 0.15,
         "cachedInputCostPerMillionTokens": 0.01,
@@ -991,7 +991,7 @@
       {
         "name": "gemini-3-pro-preview",
         "label": "Gemini 3 Pro Preview",
-        "modelFamily": "gemini",
+        "modelFamily": "GEMINI",
         "inputCostPerMillionTokens": 2,
         "outputCostPerMillionTokens": 12,
         "cachedInputCostPerMillionTokens": 0.2,
@@ -1009,7 +1009,7 @@
       {
         "name": "gemini-2.0-flash-lite",
         "label": "Gemini 2.0 Flash Lite",
-        "modelFamily": "gemini",
+        "modelFamily": "GEMINI",
         "inputCostPerMillionTokens": 0.075,
         "outputCostPerMillionTokens": 0.3,
         "contextWindowTokens": 1048576,
@@ -1019,7 +1019,7 @@
       {
         "name": "gemini-1.5-flash",
         "label": "Gemini 1.5 Flash",
-        "modelFamily": "gemini",
+        "modelFamily": "GEMINI",
         "inputCostPerMillionTokens": 0.075,
         "outputCostPerMillionTokens": 0.3,
         "cachedInputCostPerMillionTokens": 0.01875,
@@ -1030,7 +1030,7 @@
       {
         "name": "gemini-flash-lite-latest",
         "label": "Gemini Flash-Lite Latest",
-        "modelFamily": "gemini",
+        "modelFamily": "GEMINI",
         "inputCostPerMillionTokens": 0.1,
         "outputCostPerMillionTokens": 0.4,
         "cachedInputCostPerMillionTokens": 0.025,
@@ -1042,7 +1042,7 @@
       {
         "name": "gemini-2.5-pro",
         "label": "Gemini 2.5 Pro",
-        "modelFamily": "gemini",
+        "modelFamily": "GEMINI",
         "inputCostPerMillionTokens": 1.25,
         "outputCostPerMillionTokens": 10,
         "cachedInputCostPerMillionTokens": 0.31,
@@ -1054,7 +1054,7 @@
       {
         "name": "gemini-2.0-flash",
         "label": "Gemini 2.0 Flash",
-        "modelFamily": "gemini",
+        "modelFamily": "GEMINI",
         "inputCostPerMillionTokens": 0.1,
         "outputCostPerMillionTokens": 0.4,
         "cachedInputCostPerMillionTokens": 0.025,
@@ -1065,7 +1065,7 @@
       {
         "name": "gemini-1.5-pro",
         "label": "Gemini 1.5 Pro",
-        "modelFamily": "gemini",
+        "modelFamily": "GEMINI",
         "inputCostPerMillionTokens": 1.25,
         "outputCostPerMillionTokens": 5,
         "cachedInputCostPerMillionTokens": 0.3125,
@@ -1083,7 +1083,7 @@
       {
         "name": "devstral-medium-2507",
         "label": "Devstral Medium",
-        "modelFamily": "mistral",
+        "modelFamily": "MISTRAL",
         "inputCostPerMillionTokens": 0.4,
         "outputCostPerMillionTokens": 2,
         "contextWindowTokens": 128000,
@@ -1092,7 +1092,7 @@
       {
         "name": "labs-devstral-small-2512",
         "label": "Devstral Small 2",
-        "modelFamily": "mistral",
+        "modelFamily": "MISTRAL",
         "inputCostPerMillionTokens": 0,
         "outputCostPerMillionTokens": 0,
         "contextWindowTokens": 256000,
@@ -1102,7 +1102,7 @@
       {
         "name": "devstral-medium-latest",
         "label": "Devstral 2 (latest)",
-        "modelFamily": "mistral",
+        "modelFamily": "MISTRAL",
         "inputCostPerMillionTokens": 0.4,
         "outputCostPerMillionTokens": 2,
         "contextWindowTokens": 262144,
@@ -1111,7 +1111,7 @@
       {
         "name": "open-mistral-7b",
         "label": "Mistral 7B",
-        "modelFamily": "mistral",
+        "modelFamily": "MISTRAL",
         "inputCostPerMillionTokens": 0.25,
         "outputCostPerMillionTokens": 0.25,
         "contextWindowTokens": 8000,
@@ -1120,7 +1120,7 @@
       {
         "name": "mistral-small-2506",
         "label": "Mistral Small 3.2",
-        "modelFamily": "mistral",
+        "modelFamily": "MISTRAL",
         "inputCostPerMillionTokens": 0.1,
         "outputCostPerMillionTokens": 0.3,
         "contextWindowTokens": 128000,
@@ -1130,7 +1130,7 @@
       {
         "name": "mistral-medium-2505",
         "label": "Mistral Medium 3",
-        "modelFamily": "mistral",
+        "modelFamily": "MISTRAL",
         "inputCostPerMillionTokens": 0.4,
         "outputCostPerMillionTokens": 2,
         "contextWindowTokens": 131072,
@@ -1140,7 +1140,7 @@
       {
         "name": "codestral-latest",
         "label": "Codestral (latest)",
-        "modelFamily": "mistral",
+        "modelFamily": "MISTRAL",
         "inputCostPerMillionTokens": 0.3,
         "outputCostPerMillionTokens": 0.9,
         "contextWindowTokens": 256000,
@@ -1149,7 +1149,7 @@
       {
         "name": "ministral-8b-latest",
         "label": "Ministral 8B (latest)",
-        "modelFamily": "mistral",
+        "modelFamily": "MISTRAL",
         "inputCostPerMillionTokens": 0.1,
         "outputCostPerMillionTokens": 0.1,
         "contextWindowTokens": 128000,
@@ -1158,7 +1158,7 @@
       {
         "name": "magistral-small",
         "label": "Magistral Small",
-        "modelFamily": "mistral",
+        "modelFamily": "MISTRAL",
         "inputCostPerMillionTokens": 0.5,
         "outputCostPerMillionTokens": 1.5,
         "contextWindowTokens": 128000,
@@ -1168,7 +1168,7 @@
       {
         "name": "mistral-large-2512",
         "label": "Mistral Large 3",
-        "modelFamily": "mistral",
+        "modelFamily": "MISTRAL",
         "inputCostPerMillionTokens": 0.5,
         "outputCostPerMillionTokens": 1.5,
         "contextWindowTokens": 262144,
@@ -1178,7 +1178,7 @@
       {
         "name": "ministral-3b-latest",
         "label": "Ministral 3B (latest)",
-        "modelFamily": "mistral",
+        "modelFamily": "MISTRAL",
         "inputCostPerMillionTokens": 0.04,
         "outputCostPerMillionTokens": 0.04,
         "contextWindowTokens": 128000,
@@ -1187,7 +1187,7 @@
       {
         "name": "devstral-small-2505",
         "label": "Devstral Small 2505",
-        "modelFamily": "mistral",
+        "modelFamily": "MISTRAL",
         "inputCostPerMillionTokens": 0.1,
         "outputCostPerMillionTokens": 0.3,
         "contextWindowTokens": 128000,
@@ -1196,7 +1196,7 @@
       {
         "name": "pixtral-12b",
         "label": "Pixtral 12B",
-        "modelFamily": "mistral",
+        "modelFamily": "MISTRAL",
         "inputCostPerMillionTokens": 0.15,
         "outputCostPerMillionTokens": 0.15,
         "contextWindowTokens": 128000,
@@ -1206,7 +1206,7 @@
       {
         "name": "open-mixtral-8x7b",
         "label": "Mixtral 8x7B",
-        "modelFamily": "mistral",
+        "modelFamily": "MISTRAL",
         "inputCostPerMillionTokens": 0.7,
         "outputCostPerMillionTokens": 0.7,
         "contextWindowTokens": 32000,
@@ -1215,7 +1215,7 @@
       {
         "name": "pixtral-large-latest",
         "label": "Pixtral Large (latest)",
-        "modelFamily": "mistral",
+        "modelFamily": "MISTRAL",
         "inputCostPerMillionTokens": 2,
         "outputCostPerMillionTokens": 6,
         "contextWindowTokens": 128000,
@@ -1225,7 +1225,7 @@
       {
         "name": "mistral-nemo",
         "label": "Mistral Nemo",
-        "modelFamily": "mistral",
+        "modelFamily": "MISTRAL",
         "inputCostPerMillionTokens": 0.15,
         "outputCostPerMillionTokens": 0.15,
         "contextWindowTokens": 128000,
@@ -1234,7 +1234,7 @@
       {
         "name": "devstral-2512",
         "label": "Devstral 2",
-        "modelFamily": "mistral",
+        "modelFamily": "MISTRAL",
         "inputCostPerMillionTokens": 0.4,
         "outputCostPerMillionTokens": 2,
         "contextWindowTokens": 262144,
@@ -1243,7 +1243,7 @@
       {
         "name": "mistral-large-latest",
         "label": "Mistral Large (latest)",
-        "modelFamily": "mistral",
+        "modelFamily": "MISTRAL",
         "inputCostPerMillionTokens": 0.5,
         "outputCostPerMillionTokens": 1.5,
         "contextWindowTokens": 262144,
@@ -1253,7 +1253,7 @@
       {
         "name": "mistral-medium-2508",
         "label": "Mistral Medium 3.1",
-        "modelFamily": "mistral",
+        "modelFamily": "MISTRAL",
         "inputCostPerMillionTokens": 0.4,
         "outputCostPerMillionTokens": 2,
         "contextWindowTokens": 262144,
@@ -1263,7 +1263,7 @@
       {
         "name": "mistral-large-2411",
         "label": "Mistral Large 2.1",
-        "modelFamily": "mistral",
+        "modelFamily": "MISTRAL",
         "inputCostPerMillionTokens": 2,
         "outputCostPerMillionTokens": 6,
         "contextWindowTokens": 131072,
@@ -1272,7 +1272,7 @@
       {
         "name": "mistral-small-latest",
         "label": "Mistral Small (latest)",
-        "modelFamily": "mistral",
+        "modelFamily": "MISTRAL",
         "inputCostPerMillionTokens": 0.1,
         "outputCostPerMillionTokens": 0.3,
         "contextWindowTokens": 128000,
@@ -1282,7 +1282,7 @@
       {
         "name": "open-mixtral-8x22b",
         "label": "Mixtral 8x22B",
-        "modelFamily": "mistral",
+        "modelFamily": "MISTRAL",
         "inputCostPerMillionTokens": 2,
         "outputCostPerMillionTokens": 6,
         "contextWindowTokens": 64000,
@@ -1291,7 +1291,7 @@
       {
         "name": "mistral-medium-latest",
         "label": "Mistral Medium (latest)",
-        "modelFamily": "mistral",
+        "modelFamily": "MISTRAL",
         "inputCostPerMillionTokens": 0.4,
         "outputCostPerMillionTokens": 2,
         "contextWindowTokens": 128000,
@@ -1301,7 +1301,7 @@
       {
         "name": "devstral-small-2507",
         "label": "Devstral Small",
-        "modelFamily": "mistral",
+        "modelFamily": "MISTRAL",
         "inputCostPerMillionTokens": 0.1,
         "outputCostPerMillionTokens": 0.3,
         "contextWindowTokens": 128000,
@@ -1310,7 +1310,7 @@
       {
         "name": "magistral-medium-latest",
         "label": "Magistral Medium (latest)",
-        "modelFamily": "mistral",
+        "modelFamily": "MISTRAL",
         "inputCostPerMillionTokens": 2,
         "outputCostPerMillionTokens": 5,
         "contextWindowTokens": 128000,
@@ -1327,7 +1327,7 @@
       {
         "name": "grok-2-1212",
         "label": "Grok 2 (1212)",
-        "modelFamily": "grok",
+        "modelFamily": "GROK",
         "inputCostPerMillionTokens": 2,
         "outputCostPerMillionTokens": 10,
         "cachedInputCostPerMillionTokens": 2,
@@ -1337,7 +1337,7 @@
       {
         "name": "grok-2",
         "label": "Grok 2",
-        "modelFamily": "grok",
+        "modelFamily": "GROK",
         "inputCostPerMillionTokens": 2,
         "outputCostPerMillionTokens": 10,
         "cachedInputCostPerMillionTokens": 2,
@@ -1347,7 +1347,7 @@
       {
         "name": "grok-3-fast-latest",
         "label": "Grok 3 Fast Latest",
-        "modelFamily": "grok",
+        "modelFamily": "GROK",
         "inputCostPerMillionTokens": 5,
         "outputCostPerMillionTokens": 25,
         "cachedInputCostPerMillionTokens": 1.25,
@@ -1357,7 +1357,7 @@
       {
         "name": "grok-2-vision",
         "label": "Grok 2 Vision",
-        "modelFamily": "grok",
+        "modelFamily": "GROK",
         "inputCostPerMillionTokens": 2,
         "outputCostPerMillionTokens": 10,
         "cachedInputCostPerMillionTokens": 2,
@@ -1368,7 +1368,7 @@
       {
         "name": "grok-3",
         "label": "Grok 3",
-        "modelFamily": "grok",
+        "modelFamily": "GROK",
         "inputCostPerMillionTokens": 3,
         "outputCostPerMillionTokens": 15,
         "cachedInputCostPerMillionTokens": 0.75,
@@ -1378,7 +1378,7 @@
       {
         "name": "grok-code-fast-1",
         "label": "Grok Code Fast 1",
-        "modelFamily": "grok",
+        "modelFamily": "GROK",
         "inputCostPerMillionTokens": 0.2,
         "outputCostPerMillionTokens": 1.5,
         "cachedInputCostPerMillionTokens": 0.02,
@@ -1389,7 +1389,7 @@
       {
         "name": "grok-2-vision-1212",
         "label": "Grok 2 Vision (1212)",
-        "modelFamily": "grok",
+        "modelFamily": "GROK",
         "inputCostPerMillionTokens": 2,
         "outputCostPerMillionTokens": 10,
         "cachedInputCostPerMillionTokens": 2,
@@ -1400,7 +1400,7 @@
       {
         "name": "grok-4-1-fast-non-reasoning",
         "label": "Grok 4.1 Fast (Non-Reasoning)",
-        "modelFamily": "grok",
+        "modelFamily": "GROK",
         "inputCostPerMillionTokens": 0.2,
         "outputCostPerMillionTokens": 0.5,
         "cachedInputCostPerMillionTokens": 0.05,
@@ -1411,7 +1411,7 @@
       {
         "name": "grok-beta",
         "label": "Grok Beta",
-        "modelFamily": "grok",
+        "modelFamily": "GROK",
         "inputCostPerMillionTokens": 5,
         "outputCostPerMillionTokens": 15,
         "cachedInputCostPerMillionTokens": 5,
@@ -1421,7 +1421,7 @@
       {
         "name": "grok-3-mini-fast",
         "label": "Grok 3 Mini Fast",
-        "modelFamily": "grok",
+        "modelFamily": "GROK",
         "inputCostPerMillionTokens": 0.6,
         "outputCostPerMillionTokens": 4,
         "cachedInputCostPerMillionTokens": 0.15,
@@ -1432,7 +1432,7 @@
       {
         "name": "grok-4-fast",
         "label": "Grok 4 Fast",
-        "modelFamily": "grok",
+        "modelFamily": "GROK",
         "inputCostPerMillionTokens": 0.2,
         "outputCostPerMillionTokens": 0.5,
         "cachedInputCostPerMillionTokens": 0.05,
@@ -1444,7 +1444,7 @@
       {
         "name": "grok-4",
         "label": "Grok 4",
-        "modelFamily": "grok",
+        "modelFamily": "GROK",
         "inputCostPerMillionTokens": 3,
         "outputCostPerMillionTokens": 15,
         "cachedInputCostPerMillionTokens": 0.75,
@@ -1455,7 +1455,7 @@
       {
         "name": "grok-3-latest",
         "label": "Grok 3 Latest",
-        "modelFamily": "grok",
+        "modelFamily": "GROK",
         "inputCostPerMillionTokens": 3,
         "outputCostPerMillionTokens": 15,
         "cachedInputCostPerMillionTokens": 0.75,
@@ -1465,7 +1465,7 @@
       {
         "name": "grok-4-1-fast",
         "label": "Grok 4.1 Fast",
-        "modelFamily": "grok",
+        "modelFamily": "GROK",
         "inputCostPerMillionTokens": 0.2,
         "outputCostPerMillionTokens": 0.5,
         "cachedInputCostPerMillionTokens": 0.05,
@@ -1477,7 +1477,7 @@
       {
         "name": "grok-2-vision-latest",
         "label": "Grok 2 Vision Latest",
-        "modelFamily": "grok",
+        "modelFamily": "GROK",
         "inputCostPerMillionTokens": 2,
         "outputCostPerMillionTokens": 10,
         "cachedInputCostPerMillionTokens": 2,
@@ -1488,7 +1488,7 @@
       {
         "name": "grok-3-mini-latest",
         "label": "Grok 3 Mini Latest",
-        "modelFamily": "grok",
+        "modelFamily": "GROK",
         "inputCostPerMillionTokens": 0.3,
         "outputCostPerMillionTokens": 0.5,
         "cachedInputCostPerMillionTokens": 0.075,
@@ -1499,7 +1499,7 @@
       {
         "name": "grok-3-mini",
         "label": "Grok 3 Mini",
-        "modelFamily": "grok",
+        "modelFamily": "GROK",
         "inputCostPerMillionTokens": 0.3,
         "outputCostPerMillionTokens": 0.5,
         "cachedInputCostPerMillionTokens": 0.075,
@@ -1510,7 +1510,7 @@
       {
         "name": "grok-3-mini-fast-latest",
         "label": "Grok 3 Mini Fast Latest",
-        "modelFamily": "grok",
+        "modelFamily": "GROK",
         "inputCostPerMillionTokens": 0.6,
         "outputCostPerMillionTokens": 4,
         "cachedInputCostPerMillionTokens": 0.15,
@@ -1521,7 +1521,7 @@
       {
         "name": "grok-4.20-0309-reasoning",
         "label": "Grok 4.20 (Reasoning)",
-        "modelFamily": "grok",
+        "modelFamily": "GROK",
         "inputCostPerMillionTokens": 2,
         "outputCostPerMillionTokens": 6,
         "cachedInputCostPerMillionTokens": 0.2,
@@ -1539,7 +1539,7 @@
       {
         "name": "grok-2-latest",
         "label": "Grok 2 Latest",
-        "modelFamily": "grok",
+        "modelFamily": "GROK",
         "inputCostPerMillionTokens": 2,
         "outputCostPerMillionTokens": 10,
         "cachedInputCostPerMillionTokens": 2,
@@ -1549,7 +1549,7 @@
       {
         "name": "grok-4-fast-non-reasoning",
         "label": "Grok 4 Fast (Non-Reasoning)",
-        "modelFamily": "grok",
+        "modelFamily": "GROK",
         "inputCostPerMillionTokens": 0.2,
         "outputCostPerMillionTokens": 0.5,
         "cachedInputCostPerMillionTokens": 0.05,
@@ -1560,7 +1560,7 @@
       {
         "name": "grok-vision-beta",
         "label": "Grok Vision Beta",
-        "modelFamily": "grok",
+        "modelFamily": "GROK",
         "inputCostPerMillionTokens": 5,
         "outputCostPerMillionTokens": 15,
         "cachedInputCostPerMillionTokens": 5,
@@ -1571,7 +1571,7 @@
       {
         "name": "grok-4.20-0309-non-reasoning",
         "label": "Grok 4.20 (Non-Reasoning)",
-        "modelFamily": "grok",
+        "modelFamily": "GROK",
         "inputCostPerMillionTokens": 2,
         "outputCostPerMillionTokens": 6,
         "cachedInputCostPerMillionTokens": 0.2,
@@ -1588,7 +1588,7 @@
       {
         "name": "grok-3-fast",
         "label": "Grok 3 Fast",
-        "modelFamily": "grok",
+        "modelFamily": "GROK",
         "inputCostPerMillionTokens": 5,
         "outputCostPerMillionTokens": 25,
         "cachedInputCostPerMillionTokens": 1.25,
diff --git a/packages/twenty-server/src/engine/metadata-modules/ai/ai-models/constants/ai-models-types.const.spec.ts b/packages/twenty-server/src/engine/metadata-modules/ai/ai-models/constants/ai-models-types.const.spec.ts
index 48348699bdb3a..73b89a0acc250 100644
--- a/packages/twenty-server/src/engine/metadata-modules/ai/ai-models/constants/ai-models-types.const.spec.ts
+++ b/packages/twenty-server/src/engine/metadata-modules/ai/ai-models/constants/ai-models-types.const.spec.ts
@@ -8,7 +8,7 @@ import { SdkProviderFactoryService } from 'src/engine/metadata-modules/ai/ai-mod
 import { buildCompositeModelId } from 'src/engine/metadata-modules/ai/ai-models/utils/composite-model-id.util';
 import { loadDefaultAiProviders } from 'src/engine/metadata-modules/ai/ai-models/utils/load-default-ai-providers.util';
 import { type AiProvidersConfig } from 'src/engine/metadata-modules/ai/ai-models/types/ai-providers-config.type';
-import { DEFAULT_SMART_MODEL } from 'src/engine/metadata-modules/ai/ai-models/types/default-smart-model.const';
+import { AUTO_SELECT_SMART_MODEL_ID } from 'twenty-shared/constants';
 
 const DEFAULT_PROVIDERS: AiProvidersConfig = loadDefaultAiProviders();
 
@@ -125,13 +125,13 @@ describe('AiModelRegistryService', () => {
     service = module.get<AiModelRegistryService>(AiModelRegistryService);
   });
 
-  it('should throw when no models are available for DEFAULT_SMART_MODEL', () => {
-    expect(() => service.getEffectiveModelConfig(DEFAULT_SMART_MODEL)).toThrow(
+  it('should throw when no models are available for AUTO_SELECT_SMART_MODEL_ID', () => {
+    expect(() => service.getEffectiveModelConfig(AUTO_SELECT_SMART_MODEL_ID)).toThrow(
       'No AI models are available. Configure at least one AI provider.',
     );
   });
 
-  it('should return effective model config for DEFAULT_SMART_MODEL when models are available', () => {
+  it('should return effective model config for AUTO_SELECT_SMART_MODEL_ID when models are available', () => {
     jest.spyOn(service, 'getAvailableModels').mockReturnValue([
       {
         modelId: 'openai/gpt-5.2',
@@ -146,14 +146,14 @@ describe('AiModelRegistryService', () => {
       model: {} as any,
     });
 
-    const result = service.getEffectiveModelConfig(DEFAULT_SMART_MODEL);
+    const result = service.getEffectiveModelConfig(AUTO_SELECT_SMART_MODEL_ID);
 
     expect(result).toBeDefined();
     expect(result.modelId).toBe('openai/gpt-5.2');
     expect(result.sdkPackage).toBe('@ai-sdk/openai');
   });
 
-  it('should return effective model config for DEFAULT_SMART_MODEL with custom model', () => {
+  it('should return effective model config for AUTO_SELECT_SMART_MODEL_ID with custom model', () => {
     jest.spyOn(service, 'getAvailableModels').mockReturnValue([
       {
         modelId: 'custom/mistral',
@@ -168,7 +168,7 @@ describe('AiModelRegistryService', () => {
       model: {} as any,
     });
 
-    const result = service.getEffectiveModelConfig(DEFAULT_SMART_MODEL);
+    const result = service.getEffectiveModelConfig(AUTO_SELECT_SMART_MODEL_ID);
 
     expect(result).toBeDefined();
     expect(result.modelId).toBe('custom/mistral');
diff --git a/packages/twenty-server/src/engine/metadata-modules/ai/ai-models/services/ai-model-registry.service.ts b/packages/twenty-server/src/engine/metadata-modules/ai/ai-models/services/ai-model-registry.service.ts
index d3fb129e6784c..87b9fb1615d2b 100644
--- a/packages/twenty-server/src/engine/metadata-modules/ai/ai-models/services/ai-model-registry.service.ts
+++ b/packages/twenty-server/src/engine/metadata-modules/ai/ai-models/services/ai-model-registry.service.ts
@@ -17,12 +17,15 @@ import { type AiProviderConfig } from 'src/engine/metadata-modules/ai/ai-models/
 import { type AiProviderModelConfig } from 'src/engine/metadata-modules/ai/ai-models/types/ai-provider-model-config.type';
 import { type AiProvidersConfig } from 'src/engine/metadata-modules/ai/ai-models/types/ai-providers-config.type';
 import { DEFAULT_CONTEXT_WINDOW_TOKENS } from 'src/engine/metadata-modules/ai/ai-models/types/default-context-window-tokens.const';
-import { DEFAULT_FAST_MODEL } from 'src/engine/metadata-modules/ai/ai-models/types/default-fast-model.const';
+import {
+  AUTO_SELECT_FAST_MODEL_ID,
+  AUTO_SELECT_SMART_MODEL_ID,
+} from 'twenty-shared/constants';
+import { isAutoSelectModelId } from 'twenty-shared/utils';
+
 import { DEFAULT_MAX_OUTPUT_TOKENS } from 'src/engine/metadata-modules/ai/ai-models/types/default-max-output-tokens.const';
-import { DEFAULT_SMART_MODEL } from 'src/engine/metadata-modules/ai/ai-models/types/default-smart-model.const';
 import { buildCompositeModelId } from 'src/engine/metadata-modules/ai/ai-models/utils/composite-model-id.util';
 import { inferModelFamily } from 'src/engine/metadata-modules/ai/ai-models/utils/infer-model-family.util';
-import { isDefaultModelSentinel } from 'src/engine/metadata-modules/ai/ai-models/utils/is-default-model-sentinel.util';
 import {
   isModelAllowedByWorkspace,
   type WorkspaceModelAvailabilitySettings,
@@ -203,9 +206,9 @@ export class AiModelRegistryService {
   }
 
   getEffectiveModelConfig(modelId: string): AIModelConfig {
-    if (isDefaultModelSentinel(modelId)) {
+    if (isAutoSelectModelId(modelId)) {
       const defaultModel =
-        modelId === DEFAULT_FAST_MODEL
+        modelId === AUTO_SELECT_FAST_MODEL_ID
           ? this.getDefaultSpeedModel()
           : this.getDefaultPerformanceModel();
 
@@ -253,7 +256,7 @@ export class AiModelRegistryService {
   }
 
   isModelAdminAllowed(modelId: string): boolean {
-    if (isDefaultModelSentinel(modelId)) {
+    if (isAutoSelectModelId(modelId)) {
       return true;
     }
 
@@ -360,7 +363,7 @@ export class AiModelRegistryService {
 
   resolveModelForAgent(agent: { modelId: string } | null): RegisteredAIModel {
     const aiModel = this.getEffectiveModelConfig(
-      agent?.modelId ?? DEFAULT_SMART_MODEL,
+      agent?.modelId ?? AUTO_SELECT_SMART_MODEL_ID,
     );
 
     const registeredModel = this.getModel(aiModel.modelId);
diff --git a/packages/twenty-server/src/engine/metadata-modules/ai/ai-models/types/default-fast-model.const.ts b/packages/twenty-server/src/engine/metadata-modules/ai/ai-models/types/default-fast-model.const.ts
deleted file mode 100644
index 84e98b83087b1..0000000000000
--- a/packages/twenty-server/src/engine/metadata-modules/ai/ai-models/types/default-fast-model.const.ts
+++ /dev/null
@@ -1 +0,0 @@
-export const DEFAULT_FAST_MODEL = 'default-fast-model' as const;
diff --git a/packages/twenty-server/src/engine/metadata-modules/ai/ai-models/types/default-smart-model.const.ts b/packages/twenty-server/src/engine/metadata-modules/ai/ai-models/types/default-smart-model.const.ts
deleted file mode 100644
index f85d6f971916b..0000000000000
--- a/packages/twenty-server/src/engine/metadata-modules/ai/ai-models/types/default-smart-model.const.ts
+++ /dev/null
@@ -1 +0,0 @@
-export const DEFAULT_SMART_MODEL = 'default-smart-model' as const;
diff --git a/packages/twenty-server/src/engine/metadata-modules/ai/ai-models/types/model-family.enum.ts b/packages/twenty-server/src/engine/metadata-modules/ai/ai-models/types/model-family.enum.ts
index 96f04e55c05a5..01aba56a7c56a 100644
--- a/packages/twenty-server/src/engine/metadata-modules/ai/ai-models/types/model-family.enum.ts
+++ b/packages/twenty-server/src/engine/metadata-modules/ai/ai-models/types/model-family.enum.ts
@@ -1,7 +1,7 @@
 export enum ModelFamily {
-  GPT = 'gpt',
-  CLAUDE = 'claude',
-  GEMINI = 'gemini',
-  MISTRAL = 'mistral',
-  GROK = 'grok',
+  GPT = 'GPT',
+  CLAUDE = 'CLAUDE',
+  GEMINI = 'GEMINI',
+  MISTRAL = 'MISTRAL',
+  GROK = 'GROK',
 }
diff --git a/packages/twenty-server/src/engine/metadata-modules/ai/ai-models/utils/is-default-model-sentinel.util.ts b/packages/twenty-server/src/engine/metadata-modules/ai/ai-models/utils/is-default-model-sentinel.util.ts
deleted file mode 100644
index 34bd198b263cd..0000000000000
--- a/packages/twenty-server/src/engine/metadata-modules/ai/ai-models/utils/is-default-model-sentinel.util.ts
+++ /dev/null
@@ -1,5 +0,0 @@
-import { DEFAULT_FAST_MODEL } from 'src/engine/metadata-modules/ai/ai-models/types/default-fast-model.const';
-import { DEFAULT_SMART_MODEL } from 'src/engine/metadata-modules/ai/ai-models/types/default-smart-model.const';
-
-export const isDefaultModelSentinel = (modelId: string): boolean =>
-  modelId === DEFAULT_FAST_MODEL || modelId === DEFAULT_SMART_MODEL;
diff --git a/packages/twenty-server/src/engine/metadata-modules/ai/ai-models/utils/is-model-allowed.util.ts b/packages/twenty-server/src/engine/metadata-modules/ai/ai-models/utils/is-model-allowed.util.ts
index 53f5b1519b0ce..3350106786e95 100644
--- a/packages/twenty-server/src/engine/metadata-modules/ai/ai-models/utils/is-model-allowed.util.ts
+++ b/packages/twenty-server/src/engine/metadata-modules/ai/ai-models/utils/is-model-allowed.util.ts
@@ -1,4 +1,4 @@
-import { isDefaultModelSentinel } from 'src/engine/metadata-modules/ai/ai-models/utils/is-default-model-sentinel.util';
+import { isAutoSelectModelId } from 'twenty-shared/utils';
 
 export type WorkspaceModelAvailabilitySettings = {
   useRecommendedModels: boolean;
@@ -10,7 +10,7 @@ export const isModelAllowedByWorkspace = (
   workspace: WorkspaceModelAvailabilitySettings,
   recommendedModelIds?: Set<string>,
 ): boolean => {
-  if (isDefaultModelSentinel(modelId)) {
+  if (isAutoSelectModelId(modelId)) {
     return true;
   }
 
diff --git a/packages/twenty-server/src/engine/workspace-manager/twenty-standard-application/utils/agent-metadata/create-standard-flat-agent-metadata.util.ts b/packages/twenty-server/src/engine/workspace-manager/twenty-standard-application/utils/agent-metadata/create-standard-flat-agent-metadata.util.ts
index cd12cc64c103c..2415607f69275 100644
--- a/packages/twenty-server/src/engine/workspace-manager/twenty-standard-application/utils/agent-metadata/create-standard-flat-agent-metadata.util.ts
+++ b/packages/twenty-server/src/engine/workspace-manager/twenty-standard-application/utils/agent-metadata/create-standard-flat-agent-metadata.util.ts
@@ -1,4 +1,4 @@
-import { DEFAULT_SMART_MODEL } from 'src/engine/metadata-modules/ai/ai-models/types/default-smart-model.const';
+import { AUTO_SELECT_SMART_MODEL_ID } from 'twenty-shared/constants';
 import { type FlatAgent } from 'src/engine/metadata-modules/flat-agent/types/flat-agent.type';
 import { type AllStandardAgentName } from 'src/engine/workspace-manager/twenty-standard-application/types/all-standard-agent-name.type';
 import {
@@ -40,7 +40,7 @@ Response format:
 - Use markdown for readability
 
 Always base answers on official Twenty documentation. Be patient and helpful.`,
-        modelId: DEFAULT_SMART_MODEL,
+        modelId: AUTO_SELECT_SMART_MODEL_ID,
         responseFormat: { type: 'text' },
         isCustom: false,
         modelConfiguration: {},
diff --git a/packages/twenty-server/src/modules/workflow/workflow-builder/workflow-version-step/workflow-version-step-operations.workspace-service.ts b/packages/twenty-server/src/modules/workflow/workflow-builder/workflow-version-step/workflow-version-step-operations.workspace-service.ts
index efcd5e9fd9e2c..4c1de0627a5d2 100644
--- a/packages/twenty-server/src/modules/workflow/workflow-builder/workflow-version-step/workflow-version-step-operations.workspace-service.ts
+++ b/packages/twenty-server/src/modules/workflow/workflow-builder/workflow-version-step/workflow-version-step-operations.workspace-service.ts
@@ -19,7 +19,7 @@ import { getFlatFieldsFromFlatObjectMetadata } from 'src/engine/api/graphql/work
 import { type WorkflowStepPositionInput } from 'src/engine/core-modules/workflow/dtos/update-workflow-step-position.input';
 import { AiAgentRoleService } from 'src/engine/metadata-modules/ai/ai-agent-role/ai-agent-role.service';
 import { AgentService } from 'src/engine/metadata-modules/ai/ai-agent/agent.service';
-import { DEFAULT_SMART_MODEL } from 'src/engine/metadata-modules/ai/ai-models/types/default-smart-model.const';
+import { AUTO_SELECT_SMART_MODEL_ID } from 'twenty-shared/constants';
 import { WorkspaceManyOrAllFlatEntityMapsCacheService } from 'src/engine/metadata-modules/flat-entity/services/workspace-many-or-all-flat-entity-maps-cache.service';
 import { LogicFunctionFromSourceService } from 'src/engine/metadata-modules/logic-function/services/logic-function-from-source.service';
 import { findFlatLogicFunctionOrThrow } from 'src/engine/metadata-modules/logic-function/utils/find-flat-logic-function-or-throw.util';
@@ -446,7 +446,7 @@ export class WorkflowVersionStepOperationsWorkspaceService {
             description: '',
             prompt:
               'You are a helpful AI assistant. Complete the task based on the workflow context.',
-            modelId: DEFAULT_SMART_MODEL,
+            modelId: AUTO_SELECT_SMART_MODEL_ID,
             responseFormat: { type: 'text' },
             isCustom: true,
           },
diff --git a/packages/twenty-server/src/modules/workflow/workflow-executor/workflow-actions/ai-agent/ai-agent.workflow-action.ts b/packages/twenty-server/src/modules/workflow/workflow-executor/workflow-actions/ai-agent/ai-agent.workflow-action.ts
index de3f41e799634..c9d76a0aee412 100644
--- a/packages/twenty-server/src/modules/workflow/workflow-executor/workflow-actions/ai-agent/ai-agent.workflow-action.ts
+++ b/packages/twenty-server/src/modules/workflow/workflow-executor/workflow-actions/ai-agent/ai-agent.workflow-action.ts
@@ -9,7 +9,7 @@ import { type WorkflowAction } from 'src/modules/workflow/workflow-executor/inte
 import { AgentAsyncExecutorService } from 'src/engine/metadata-modules/ai/ai-agent-execution/services/agent-async-executor.service';
 import { AgentEntity } from 'src/engine/metadata-modules/ai/ai-agent/entities/agent.entity';
 import { AiBillingService } from 'src/engine/metadata-modules/ai/ai-billing/services/ai-billing.service';
-import { DEFAULT_SMART_MODEL } from 'src/engine/metadata-modules/ai/ai-models/types/default-smart-model.const';
+import { AUTO_SELECT_SMART_MODEL_ID } from 'twenty-shared/constants';
 import {
   WorkflowStepExecutorException,
   WorkflowStepExecutorExceptionCode,
@@ -90,7 +90,7 @@ export class AiAgentWorkflowAction implements WorkflowAction {
       });
 
     await this.aiBillingService.calculateAndBillUsage(
-      agent?.modelId ?? DEFAULT_SMART_MODEL,
+      agent?.modelId ?? AUTO_SELECT_SMART_MODEL_ID,
       { usage, cacheCreationTokens },
       workspaceId,
       agent?.id || null,
diff --git a/packages/twenty-shared/src/constants/AutoSelectFastModelId.ts b/packages/twenty-shared/src/constants/AutoSelectFastModelId.ts
new file mode 100644
index 0000000000000..5a18f5f460229
--- /dev/null
+++ b/packages/twenty-shared/src/constants/AutoSelectFastModelId.ts
@@ -0,0 +1 @@
+export const AUTO_SELECT_FAST_MODEL_ID = 'default-fast-model' as const;
diff --git a/packages/twenty-shared/src/constants/AutoSelectSmartModelId.ts b/packages/twenty-shared/src/constants/AutoSelectSmartModelId.ts
new file mode 100644
index 0000000000000..cc5e9ebbc4298
--- /dev/null
+++ b/packages/twenty-shared/src/constants/AutoSelectSmartModelId.ts
@@ -0,0 +1 @@
+export const AUTO_SELECT_SMART_MODEL_ID = 'default-smart-model' as const;
diff --git a/packages/twenty-shared/src/constants/index.ts b/packages/twenty-shared/src/constants/index.ts
index 3b06cd9ef7ca4..ea586dd07de17 100644
--- a/packages/twenty-shared/src/constants/index.ts
+++ b/packages/twenty-shared/src/constants/index.ts
@@ -7,6 +7,8 @@
  *                              |___/
  */
 
+export { AUTO_SELECT_FAST_MODEL_ID } from './AutoSelectFastModelId';
+export { AUTO_SELECT_SMART_MODEL_ID } from './AutoSelectSmartModelId';
 export { BACKEND_BATCH_REQUEST_MAX_COUNT } from './BackendBatchRequestMaxCount';
 export { CalendarStartDay } from './CalendarStartDay';
 export { COMPOSITE_FIELD_TYPE_SUB_FIELDS_NAMES } from './CompositeFieldTypeSubFieldsNames';
diff --git a/packages/twenty-shared/src/utils/index.ts b/packages/twenty-shared/src/utils/index.ts
index ebdf8072ed263..01517cad0b166 100644
--- a/packages/twenty-shared/src/utils/index.ts
+++ b/packages/twenty-shared/src/utils/index.ts
@@ -143,6 +143,7 @@ export {
   getLogoUrlFromDomainName,
 } from './image/getLogoUrlFromDomainName';
 export { getUniqueConstraintsFields } from './indexMetadata/getUniqueConstraintsFields';
+export { isAutoSelectModelId } from './isAutoSelectModelId';
 export { fastDeepEqual } from './json/fast-deep-equal';
 export { getAppPath } from './navigation/getAppPath';
 export { getSettingsPath } from './navigation/getSettingsPath';
diff --git a/packages/twenty-shared/src/utils/isAutoSelectModelId.ts b/packages/twenty-shared/src/utils/isAutoSelectModelId.ts
new file mode 100644
index 0000000000000..06cbf35edbe7c
--- /dev/null
+++ b/packages/twenty-shared/src/utils/isAutoSelectModelId.ts
@@ -0,0 +1,6 @@
+import { AUTO_SELECT_FAST_MODEL_ID } from '../constants/AutoSelectFastModelId';
+import { AUTO_SELECT_SMART_MODEL_ID } from '../constants/AutoSelectSmartModelId';
+
+export const isAutoSelectModelId = (modelId: string): boolean =>
+  modelId === AUTO_SELECT_FAST_MODEL_ID ||
+  modelId === AUTO_SELECT_SMART_MODEL_ID;
diff --git a/packages/twenty-ui/src/input/button/components/RoundedIconButton.tsx b/packages/twenty-ui/src/input/button/components/RoundedIconButton.tsx
index 07e86e1e656a8..a057afca7aefe 100644
--- a/packages/twenty-ui/src/input/button/components/RoundedIconButton.tsx
+++ b/packages/twenty-ui/src/input/button/components/RoundedIconButton.tsx
@@ -23,6 +23,10 @@ const StyledIconButton = styled.button<{
     color 0.1s ease-in-out,
     background 0.1s ease-in-out;
 
+  &:hover:not(:disabled) {
+    background: ${themeCssVariables.color.blue10};
+  }
+
   &:disabled {
     background: ${themeCssVariables.background.quaternary};
     color: ${themeCssVariables.font.color.tertiary};
diff --git a/packages/twenty-ui/src/input/types/SelectOption.ts b/packages/twenty-ui/src/input/types/SelectOption.ts
index 949eae87712bd..3b3f9d39f2dbf 100644
--- a/packages/twenty-ui/src/input/types/SelectOption.ts
+++ b/packages/twenty-ui/src/input/types/SelectOption.ts
@@ -10,4 +10,5 @@ export type SelectOption<
   value: Value;
   disabled?: boolean;
   color?: ThemeColor | 'transparent';
+  contextualText?: string;
 };