Merge pull request #1192 from RooVetGit/cte/move-model-fetchers

Move & clean up model fetchers

Author: cte

src/api/providers/glama.ts

@@ -1,10 +1,12 @@
 import { Anthropic } from "@anthropic-ai/sdk"
 import axios from "axios"
 import OpenAI from "openai"
-import { ApiHandler, SingleCompletionHandler } from "../"
+
 import { ApiHandlerOptions, ModelInfo, glamaDefaultModelId, glamaDefaultModelInfo } from "../../shared/api"
+import { parseApiPrice } from "../../utils/cost"
 import { convertToOpenAiMessages } from "../transform/openai-format"
 import { ApiStream } from "../transform/stream"
+import { ApiHandler, SingleCompletionHandler } from "../"
 
 const GLAMA_DEFAULT_TEMPERATURE = 0
 
@@ -190,3 +192,44 @@ export class GlamaHandler implements ApiHandler, SingleCompletionHandler {
 		}
 	}
 }
+
+export async function getGlamaModels() {
+	const models: Record<string, ModelInfo> = {}
+
+	try {
+		const response = await axios.get("https://glama.ai/api/gateway/v1/models")
+		const rawModels = response.data
+
+		for (const rawModel of rawModels) {
+			const modelInfo: ModelInfo = {
+				maxTokens: rawModel.maxTokensOutput,
+				contextWindow: rawModel.maxTokensInput,
+				supportsImages: rawModel.capabilities?.includes("input:image"),
+				supportsComputerUse: rawModel.capabilities?.includes("computer_use"),
+				supportsPromptCache: rawModel.capabilities?.includes("caching"),
+				inputPrice: parseApiPrice(rawModel.pricePerToken?.input),
+				outputPrice: parseApiPrice(rawModel.pricePerToken?.output),
+				description: undefined,
+				cacheWritesPrice: parseApiPrice(rawModel.pricePerToken?.cacheWrite),
+				cacheReadsPrice: parseApiPrice(rawModel.pricePerToken?.cacheRead),
+			}
+
+			switch (rawModel.id) {
+				case rawModel.id.startsWith("anthropic/claude-3-7-sonnet"):
+					modelInfo.maxTokens = 16384
+					break
+				case rawModel.id.startsWith("anthropic/"):
+					modelInfo.maxTokens = 8192
+					break
+				default:
+					break
+			}
+
+			models[rawModel.id] = modelInfo
+		}
+	} catch (error) {
+		console.error(`Error fetching Glama models: ${JSON.stringify(error, Object.getOwnPropertyNames(error), 2)}`)
+	}
+
+	return models
+}

src/api/providers/lmstudio.ts

@@ -1,5 +1,7 @@
 import { Anthropic } from "@anthropic-ai/sdk"
 import OpenAI from "openai"
+import axios from "axios"
+
 import { ApiHandler, SingleCompletionHandler } from "../"
 import { ApiHandlerOptions, ModelInfo, openAiModelInfoSaneDefaults } from "../../shared/api"
 import { convertToOpenAiMessages } from "../transform/openai-format"
@@ -72,3 +74,17 @@ export class LmStudioHandler implements ApiHandler, SingleCompletionHandler {
 		}
 	}
 }
+
+export async function getLmStudioModels(baseUrl = "http://localhost:1234") {
+	try {
+		if (!URL.canParse(baseUrl)) {
+			return []
+		}
+
+		const response = await axios.get(`${baseUrl}/v1/models`)
+		const modelsArray = response.data?.data?.map((model: any) => model.id) || []
+		return [...new Set<string>(modelsArray)]
+	} catch (error) {
+		return []
+	}
+}

src/api/providers/ollama.ts

@@ -1,5 +1,7 @@
 import { Anthropic } from "@anthropic-ai/sdk"
 import OpenAI from "openai"
+import axios from "axios"
+
 import { ApiHandler, SingleCompletionHandler } from "../"
 import { ApiHandlerOptions, ModelInfo, openAiModelInfoSaneDefaults } from "../../shared/api"
 import { convertToOpenAiMessages } from "../transform/openai-format"
@@ -88,3 +90,17 @@ export class OllamaHandler implements ApiHandler, SingleCompletionHandler {
 		}
 	}
 }
+
+export async function getOllamaModels(baseUrl = "http://localhost:11434") {
+	try {
+		if (!URL.canParse(baseUrl)) {
+			return []
+		}
+
+		const response = await axios.get(`${baseUrl}/api/tags`)
+		const modelsArray = response.data?.models?.map((model: any) => model.name) || []
+		return [...new Set<string>(modelsArray)]
+	} catch (error) {
+		return []
+	}
+}

src/api/providers/openai.ts

@@ -1,5 +1,6 @@
 import { Anthropic } from "@anthropic-ai/sdk"
 import OpenAI, { AzureOpenAI } from "openai"
+import axios from "axios"
 
 import {
 	ApiHandlerOptions,
@@ -166,3 +167,27 @@ export class OpenAiHandler implements ApiHandler, SingleCompletionHandler {
 		}
 	}
 }
+
+export async function getOpenAiModels(baseUrl?: string, apiKey?: string) {
+	try {
+		if (!baseUrl) {
+			return []
+		}
+
+		if (!URL.canParse(baseUrl)) {
+			return []
+		}
+
+		const config: Record<string, any> = {}
+
+		if (apiKey) {
+			config["headers"] = { Authorization: `Bearer ${apiKey}` }
+		}
+
+		const response = await axios.get(`${baseUrl}/models`, config)
+		const modelsArray = response.data?.data?.map((model: any) => model.id) || []
+		return [...new Set<string>(modelsArray)]
+	} catch (error) {
+		return []
+	}
+}

src/api/providers/openrouter.ts

@@ -1,29 +1,29 @@
 import { Anthropic } from "@anthropic-ai/sdk"
 import axios from "axios"
 import OpenAI from "openai"
-import { ApiHandler } from "../"
+import delay from "delay"
+
 import { ApiHandlerOptions, ModelInfo, openRouterDefaultModelId, openRouterDefaultModelInfo } from "../../shared/api"
+import { parseApiPrice } from "../../utils/cost"
 import { convertToOpenAiMessages } from "../transform/openai-format"
 import { ApiStreamChunk, ApiStreamUsageChunk } from "../transform/stream"
-import delay from "delay"
+import { convertToR1Format } from "../transform/r1-format"
 import { DEEP_SEEK_DEFAULT_TEMPERATURE } from "./openai"
+import { ApiHandler, SingleCompletionHandler } from ".."
 
 const OPENROUTER_DEFAULT_TEMPERATURE = 0
 
-// Add custom interface for OpenRouter params
+// Add custom interface for OpenRouter params.
 type OpenRouterChatCompletionParams = OpenAI.Chat.ChatCompletionCreateParams & {
 	transforms?: string[]
 	include_reasoning?: boolean
 }
 
-// Add custom interface for OpenRouter usage chunk
+// Add custom interface for OpenRouter usage chunk.
 interface OpenRouterApiStreamUsageChunk extends ApiStreamUsageChunk {
 	fullResponseText: string
 }
 
-import { SingleCompletionHandler } from ".."
-import { convertToR1Format } from "../transform/r1-format"
-
 export class OpenRouterHandler implements ApiHandler, SingleCompletionHandler {
 	private options: ApiHandlerOptions
 	private client: OpenAI
@@ -222,3 +222,75 @@ export class OpenRouterHandler implements ApiHandler, SingleCompletionHandler {
 		}
 	}
 }
+
+export async function getOpenRouterModels() {
+	const models: Record<string, ModelInfo> = {}
+
+	try {
+		const response = await axios.get("https://openrouter.ai/api/v1/models")
+		const rawModels = response.data.data
+
+		for (const rawModel of rawModels) {
+			const modelInfo: ModelInfo = {
+				maxTokens: rawModel.top_provider?.max_completion_tokens,
+				contextWindow: rawModel.context_length,
+				supportsImages: rawModel.architecture?.modality?.includes("image"),
+				supportsPromptCache: false,
+				inputPrice: parseApiPrice(rawModel.pricing?.prompt),
+				outputPrice: parseApiPrice(rawModel.pricing?.completion),
+				description: rawModel.description,
+			}
+
+			// NOTE: this needs to be synced with api.ts/openrouter default model info.
+			switch (true) {
+				case rawModel.id.startsWith("anthropic/claude-3.7-sonnet"):
+					modelInfo.supportsComputerUse = true
+					modelInfo.supportsPromptCache = true
+					modelInfo.cacheWritesPrice = 3.75
+					modelInfo.cacheReadsPrice = 0.3
+					modelInfo.maxTokens = 16384
+					break
+				case rawModel.id.startsWith("anthropic/claude-3.5-sonnet-20240620"):
+					modelInfo.supportsPromptCache = true
+					modelInfo.cacheWritesPrice = 3.75
+					modelInfo.cacheReadsPrice = 0.3
+					modelInfo.maxTokens = 8192
+					break
+				case rawModel.id.startsWith("anthropic/claude-3.5-sonnet"):
+					modelInfo.supportsComputerUse = true
+					modelInfo.supportsPromptCache = true
+					modelInfo.cacheWritesPrice = 3.75
+					modelInfo.cacheReadsPrice = 0.3
+					modelInfo.maxTokens = 8192
+					break
+				case rawModel.id.startsWith("anthropic/claude-3-5-haiku"):
+					modelInfo.supportsPromptCache = true
+					modelInfo.cacheWritesPrice = 1.25
+					modelInfo.cacheReadsPrice = 0.1
+					modelInfo.maxTokens = 8192
+					break
+				case rawModel.id.startsWith("anthropic/claude-3-opus"):
+					modelInfo.supportsPromptCache = true
+					modelInfo.cacheWritesPrice = 18.75
+					modelInfo.cacheReadsPrice = 1.5
+					modelInfo.maxTokens = 8192
+					break
+				case rawModel.id.startsWith("anthropic/claude-3-haiku"):
+				default:
+					modelInfo.supportsPromptCache = true
+					modelInfo.cacheWritesPrice = 0.3
+					modelInfo.cacheReadsPrice = 0.03
+					modelInfo.maxTokens = 8192
+					break
+			}
+
+			models[rawModel.id] = modelInfo
+		}
+	} catch (error) {
+		console.error(
+			`Error fetching OpenRouter models: ${JSON.stringify(error, Object.getOwnPropertyNames(error), 2)}`,
+		)
+	}
+
+	return models
+}

src/api/providers/requesty.ts

@@ -1,6 +1,9 @@
-import { OpenAiHandler, OpenAiHandlerOptions } from "./openai"
+import axios from "axios"
+
 import { ModelInfo, requestyModelInfoSaneDefaults, requestyDefaultModelId } from "../../shared/api"
-import { ApiStream, ApiStreamUsageChunk } from "../transform/stream"
+import { parseApiPrice } from "../../utils/cost"
+import { ApiStreamUsageChunk } from "../transform/stream"
+import { OpenAiHandler, OpenAiHandlerOptions } from "./openai"
 
 export class RequestyHandler extends OpenAiHandler {
 	constructor(options: OpenAiHandlerOptions) {
@@ -38,3 +41,51 @@ export class RequestyHandler extends OpenAiHandler {
 		}
 	}
 }
+
+export async function getRequestyModels({ apiKey }: { apiKey?: string }) {
+	const models: Record<string, ModelInfo> = {}
+
+	if (!apiKey) {
+		return models
+	}
+
+	try {
+		const config: Record<string, any> = {}
+		config["headers"] = { Authorization: `Bearer ${apiKey}` }
+
+		const response = await axios.get("https://router.requesty.ai/v1/models", config)
+		const rawModels = response.data.data
+
+		for (const rawModel of rawModels) {
+			const modelInfo: ModelInfo = {
+				maxTokens: rawModel.max_output_tokens,
+				contextWindow: rawModel.context_window,
+				supportsImages: rawModel.support_image,
+				supportsComputerUse: rawModel.support_computer_use,
+				supportsPromptCache: rawModel.supports_caching,
+				inputPrice: parseApiPrice(rawModel.input_price),
+				outputPrice: parseApiPrice(rawModel.output_price),
+				description: rawModel.description,
+				cacheWritesPrice: parseApiPrice(rawModel.caching_price),
+				cacheReadsPrice: parseApiPrice(rawModel.cached_price),
+			}
+
+			switch (rawModel.id) {
+				case rawModel.id.startsWith("anthropic/claude-3-7-sonnet"):
+					modelInfo.maxTokens = 16384
+					break
+				case rawModel.id.startsWith("anthropic/"):
+					modelInfo.maxTokens = 8192
+					break
+				default:
+					break
+			}
+
+			models[rawModel.id] = modelInfo
+		}
+	} catch (error) {
+		console.error(`Error fetching Requesty models: ${JSON.stringify(error, Object.getOwnPropertyNames(error), 2)}`)
+	}
+
+	return models
+}

src/api/providers/unbound.ts

@@ -1,9 +1,11 @@
 import { Anthropic } from "@anthropic-ai/sdk"
+import axios from "axios"
 import OpenAI from "openai"
-import { ApiHandler, SingleCompletionHandler } from "../"
+
 import { ApiHandlerOptions, ModelInfo, unboundDefaultModelId, unboundDefaultModelInfo } from "../../shared/api"
 import { convertToOpenAiMessages } from "../transform/openai-format"
 import { ApiStream, ApiStreamUsageChunk } from "../transform/stream"
+import { ApiHandler, SingleCompletionHandler } from "../"
 
 interface UnboundUsage extends OpenAI.CompletionUsage {
 	cache_creation_input_tokens?: number
@@ -163,3 +165,46 @@ export class UnboundHandler implements ApiHandler, SingleCompletionHandler {
 		}
 	}
 }
+
+export async function getUnboundModels() {
+	const models: Record<string, ModelInfo> = {}
+
+	try {
+		const response = await axios.get("https://api.getunbound.ai/models")
+
+		if (response.data) {
+			const rawModels: Record<string, any> = response.data
+
+			for (const [modelId, model] of Object.entries(rawModels)) {
+				const modelInfo: ModelInfo = {
+					maxTokens: model?.maxTokens ? parseInt(model.maxTokens) : undefined,
+					contextWindow: model?.contextWindow ? parseInt(model.contextWindow) : 0,
+					supportsImages: model?.supportsImages ?? false,
+					supportsPromptCache: model?.supportsPromptCaching ?? false,
+					supportsComputerUse: model?.supportsComputerUse ?? false,
+					inputPrice: model?.inputTokenPrice ? parseFloat(model.inputTokenPrice) : undefined,
+					outputPrice: model?.outputTokenPrice ? parseFloat(model.outputTokenPrice) : undefined,
+					cacheWritesPrice: model?.cacheWritePrice ? parseFloat(model.cacheWritePrice) : undefined,
+					cacheReadsPrice: model?.cacheReadPrice ? parseFloat(model.cacheReadPrice) : undefined,
+				}
+
+				switch (true) {
+					case modelId.startsWith("anthropic/claude-3-7-sonnet"):
+						modelInfo.maxTokens = 16384
+						break
+					case modelId.startsWith("anthropic/"):
+						modelInfo.maxTokens = 8192
+						break
+					default:
+						break
+				}
+
+				models[modelId] = modelInfo
+			}
+		}
+	} catch (error) {
+		console.error(`Error fetching Unbound models: ${JSON.stringify(error, Object.getOwnPropertyNames(error), 2)}`)
+	}
+
+	return models
+}