Add changelog for 2025-09-29 (#716)

Author: quantstruct-bot

fern/changelog/2025-08-30.mdx

@@ -0,0 +1,17 @@
+# Enhanced Authentication & Custom Credentials
+
+1. **Custom Credential System**: You can now create and manage custom authentication credentials using the new [`CustomCredential`](https://api.vapi.ai/api#:~:text=CustomCredential) system. This powerful new feature supports multiple authentication methods:
+    - **OAuth2 RFC 6749**: Full OAuth2 implementation for secure third-party integrations
+    - **HMAC Signing**: Cryptographic message authentication for enhanced security
+    - **Bearer Token**: Simple token-based authentication for API access
+
+2. **Bearer Authentication Plans**: Implement secure token-based authentication with [`BearerAuthenticationPlan`](https://api.vapi.ai/api#:~:text=BearerAuthenticationPlan). Key features include:
+    - `token`: Your secure bearer token value
+    - `headerName`: Custom header name (defaults to 'Authorization')
+    - `bearerPrefixEnabled`: Toggle 'Bearer ' prefix inclusion (defaults to true)
+
+3. **Enhanced Webhook Credentials**: Webhook integrations now support advanced authentication through [`WebhookCredential.authenticationPlan`](https://api.vapi.ai/api#:~:text=WebhookCredential.authenticationPlan), enabling secure webhook communications with OAuth2, HMAC, or Bearer authentication.
+
+4. **Server Authentication**: Secure your server endpoints with credential-based authentication using [`Server.credentialId`](https://api.vapi.ai/api#:~:text=Server.credentialId) to link your custom credentials to webhook destinations.
+
+5. **Tool Authentication Integration**: API request tools can now use custom credentials for secure external API calls via [`ApiRequestTool.credentialId`](https://api.vapi.ai/api#:~:text=ApiRequestTool.credentialId), eliminating the need to embed sensitive authentication details directly in tool configurations.

fern/changelog/2025-09-02.mdx

@@ -0,0 +1,36 @@
+# Recording Consent & Compliance Management
+
+1. **Recording Consent Plans**: Ensure legal compliance with call recording regulations using the new [`CompliancePlan.recordingConsentPlan`](https://api.vapi.ai/api#:~:text=CompliancePlan.recordingConsentPlan). This feature helps you meet GDPR, CCPA, and other privacy regulations by properly obtaining user consent before recording calls.
+
+2. **Verbal Consent Collection**: Implement active consent collection with [`RecordingConsentPlanVerbal`](https://api.vapi.ai/api#:~:text=RecordingConsentPlanVerbal) where users explicitly agree or decline recording:
+    - `message`: Custom consent message (e.g., "This call may be recorded for quality purposes. Say 'I agree' to consent.")
+    - `voice`: Optional dedicated voice for consent messages for better user experience
+    - `declineTool`: Execute specific tools when users decline consent
+    - `declineToolId`: Reference existing tools for decline handling
+
+3. **Stay-on-Line Consent**: Use passive consent collection with [`RecordingConsentPlanStayOnLine`](https://api.vapi.ai/api#:~:text=RecordingConsentPlanStayOnLine) where staying on the call implies consent:
+    - `message`: Informational message about recording (e.g., "For quality purposes, this call may be recorded. Please hang up if you do not consent.")
+    - `waitSeconds`: Configurable wait time (1-6 seconds) before proceeding
+    - `voice`: Optional separate voice for consent announcements
+
+4. **Recording Consent Tracking**: Monitor consent status throughout the call lifecycle with [`Call.compliance.recordingConsent`](https://api.vapi.ai/api#:~:text=Call.compliance.recordingConsent):
+    - `type`: The type of consent obtained
+    - `grantedAt`: Timestamp when consent was granted (null if not granted)
+
+5. **Enhanced End-of-Call Reports**: Recording consent information is now included in [`ServerMessageEndOfCallReport.compliance`](https://api.vapi.ai/api#:~:text=ServerMessageEndOfCallReport.compliance), providing complete compliance audit trails for your records.
+
+## Compliance Features
+<CardGroup cols={2}>
+  <Card title="Legal Compliance" icon="gavel">
+    Meet GDPR, CCPA, and other privacy regulations with built-in consent management and audit trails.
+  </Card>
+  <Card title="Flexible Consent Methods" icon="microphone">
+    Choose between verbal consent requiring explicit agreement or stay-on-line consent with implied agreement.
+  </Card>
+  <Card title="Custom Messaging" icon="message-circle">
+    Customize consent messages to match your brand voice and legal requirements with up to 1000 characters.
+  </Card>
+  <Card title="Audit Trails" icon="file-text">
+    Complete compliance records with timestamps and consent status in call artifacts and end-of-call reports.
+  </Card>
+</CardGroup>

fern/changelog/2025-09-05.mdx

@@ -0,0 +1,47 @@
+# Evaluation System Foundation
+
+1. **Evaluation Framework**: You can now systematically test your Vapi voice assistants with the new [`Eval`](https://api.vapi.ai/api#:~:text=Eval) system. Create comprehensive test scenarios to validate assistant behavior, conversation flow, and tool usage through mock conversations.
+
+2. **Mock Conversation Builder**: Design test conversations using [`Eval.messages`](https://api.vapi.ai/api#:~:text=Eval.messages) with support for multiple message types:
+    - [`ChatEvalUserMessageMock`](https://api.vapi.ai/api#:~:text=ChatEvalUserMessageMock): Simulate user inputs and questions
+    - [`ChatEvalSystemMessageMock`](https://api.vapi.ai/api#:~:text=ChatEvalSystemMessageMock): Inject system messages mid-conversation
+    - [`ChatEvalToolResponseMessageMock`](https://api.vapi.ai/api#:~:text=ChatEvalToolResponseMessageMock): Mock tool responses for consistent testing
+    - [`ChatEvalAssistantMessageEvaluation`](https://api.vapi.ai/api#:~:text=ChatEvalAssistantMessageEvaluation): Define evaluation checkpoints
+
+3. **Evaluation Types**: Currently focused on `chat.mockConversation` type evaluations, with the framework designed to support additional evaluation methods in future releases.
+
+4. **Evaluation Management**: Organize your tests with [`CreateEvalDTO`](https://api.vapi.ai/api#:~:text=CreateEvalDTO) and [`UpdateEvalDTO`](https://api.vapi.ai/api#:~:text=UpdateEvalDTO):
+    - `name`: Descriptive names up to 80 characters (e.g., "Customer Support Flow Validation")
+    - `description`: Detailed descriptions up to 500 characters explaining the test purpose
+    - `messages`: The complete mock conversation flow
+
+5. **Evaluation Endpoints**: Access your evaluations through the new [`/eval`](https://api.vapi.ai/api#:~:text=/eval) endpoint family:
+    - `GET /eval`: List all evaluations with pagination support
+    - `POST /eval`: Create new evaluations
+    - `GET /eval/{id}`: Retrieve specific evaluation details
+    - `PUT /eval/{id}`: Update existing evaluations
+
+6. **Judge Plan Architecture**: Define how assistant responses are validated using [`AssistantMessageJudgePlan`](https://api.vapi.ai/api#:~:text=AssistantMessageJudgePlan) with three evaluation methods:
+    - **Exact Match**: [`AssistantMessageJudgePlanExact`](https://api.vapi.ai/api#:~:text=AssistantMessageJudgePlanExact) for precise content and tool call validation
+    - **Regex Pattern**: [`AssistantMessageJudgePlanRegex`](https://api.vapi.ai/api#:~:text=AssistantMessageJudgePlanRegex) for flexible pattern-based evaluation
+    - **AI Judge**: [`AssistantMessageJudgePlanAI`](https://api.vapi.ai/api#:~:text=AssistantMessageJudgePlanAI) for intelligent evaluation using LLM-as-a-judge
+
+<Info>
+    This is the foundation release for the evaluation system. Evaluation execution and results processing will be available in upcoming releases. Start designing your test scenarios now to be ready for full evaluation capabilities.
+</Info>
+
+## Testing Capabilities
+<CardGroup cols={2}>
+  <Card title="Mock Conversations" icon="message-square">
+    Create realistic test scenarios with user messages, system prompts, and expected assistant responses for comprehensive flow validation.
+  </Card>
+  <Card title="Tool Call Testing" icon="wrench">
+    Validate that your assistant calls the right tools with correct parameters using <code>ChatEvalAssistantMessageMockToolCall</code>.
+  </Card>
+  <Card title="Flexible Validation" icon="check-circle">
+    Choose from exact matching, regex patterns, or AI-powered evaluation to suit different testing needs and complexity levels.
+  </Card>
+  <Card title="Evaluation Organization" icon="folder">
+    Organize tests with descriptive names and detailed documentation to maintain clear testing workflows across your team.
+  </Card>
+</CardGroup>

fern/changelog/2025-09-08.mdx

@@ -0,0 +1,44 @@
+# Enhanced Transcription Features & Speech Processing
+
+1. **Gladia Transcription Enhancements**: Improve transcription accuracy and performance with new [`GladiaTranscriber`](https://api.vapi.ai/api#:~:text=GladiaTranscriber) features:
+    - `region`: Choose between `us-west` and `eu-west` for optimal latency and data residency compliance
+    - `receivePartialTranscripts`: Enable low-latency streaming transcription for real-time conversation flow
+    - Enhanced language detection with support for both single and multiple language modes
+
+2. **Advanced Deepgram Controls**: Fine-tune speech recognition with enhanced [`DeepgramTranscriber`](https://api.vapi.ai/api#:~:text=DeepgramTranscriber) settings:
+    - `eotThreshold`: End-of-turn detection threshold for precise conversation boundaries (e.g., 0.7)
+    - `eotTimeoutMs`: Maximum wait time for end-of-turn detection in milliseconds (e.g., 5000ms)
+    - `eagerEotThreshold`: Early end-of-turn detection for responsive conversations (e.g., 0.3)
+
+3. **AssemblyAI Keyterms Enhancement**: Boost recognition accuracy for critical terms with [`AssemblyAITranscriber.keytermsPrompt`](https://api.vapi.ai/api#:~:text=AssemblyAITranscriber.keytermsPrompt):
+    - Support for up to 100 keyterms, each up to 50 characters
+    - Improved recognition for specific words and phrases
+    - Additional cost: $0.04/hour when enabled
+
+4. **Speechmatics Custom Vocabulary**: Enhance recognition accuracy with [`SpeechmaticsCustomVocabularyItem`](https://api.vapi.ai/api#:~:text=SpeechmaticsCustomVocabularyItem):
+    - `content`: The word or phrase to add (e.g., "Speechmatics")
+    - `soundsLike`: Alternative phonetic representations (e.g., ["speech mattix"]) for better pronunciation handling
+
+5. **Word-Level Confidence**: Access detailed transcription confidence data with [`CustomLLMModel.wordLevelConfidenceEnabled`](https://api.vapi.ai/api#:~:text=CustomLLMModel.wordLevelConfidenceEnabled), providing word-by-word accuracy metrics for quality assessment and debugging.
+
+6. **Enhanced Message Metadata**: Store transcription confidence and other metadata in [`UserMessage.metadata`](https://api.vapi.ai/api#:~:text=UserMessage.metadata), enabling detailed analysis of transcription quality and user speech patterns.
+
+<Warning>
+    `AssemblyAITranscriber.wordFinalizationMaxWaitTime` is now deprecated. Use the new smart endpointing plans for better speech timing control. The deprecated property will be removed in a future release.
+</Warning>
+
+## Transcription Improvements
+<CardGroup cols={2}>
+  <Card title="Regional Processing" icon="globe">
+    Choose optimal transcription regions with Gladia's <code>us-west</code> and <code>eu-west</code> options for reduced latency and compliance.
+  </Card>
+  <Card title="Real-Time Streaming" icon="zap">
+    Enable partial transcripts for immediate response processing, reducing perceived latency in conversations.
+  </Card>
+  <Card title="Advanced Speech Detection" icon="mic">
+    Fine-tune end-of-turn detection with configurable thresholds and timeouts for natural conversation flow.
+  </Card>
+  <Card title="Custom Vocabulary" icon="book">
+    Improve accuracy for domain-specific terms, company names, and technical jargon with enhanced vocabulary support.
+  </Card>
+</CardGroup>

fern/changelog/2025-09-11.mdx

@@ -0,0 +1,37 @@
+# Voice Enhancements & Minimax Improvements
+
+1. **Minimax Voice Language Support**: Enhance multilingual conversations with [`MinimaxVoice.languageBoost`](https://api.vapi.ai/api#:~:text=MinimaxVoice.languageBoost). Support for 40+ languages including:
+    - `Chinese` and `Chinese,Yue` for Mandarin and Cantonese
+    - `English`, `Spanish`, `French`, `German`, `Japanese`, `Korean`
+    - Regional variants and specialized languages like `Arabic`, `Hindi`, `Thai`
+    - `auto` mode for automatic language detection
+
+2. **Text Normalization**: Improve number reading and formatting with [`MinimaxVoice.textNormalizationEnabled`](https://api.vapi.ai/api#:~:text=MinimaxVoice.textNormalizationEnabled). When enabled, spoken numbers, dates, and formatted text are properly pronounced for natural-sounding conversations.
+
+3. **Enhanced Voice Caching**: Voice responses are now cached by default with [`MinimaxVoice.cachingEnabled`](https://api.vapi.ai/api#:~:text=MinimaxVoice.cachingEnabled) set to `true`, reducing latency for repeated phrases and improving overall conversation performance.
+
+4. **Fallback Voice Configuration**: Ensure conversation continuity with [`FallbackMinimaxVoice`](https://api.vapi.ai/api#:~:text=FallbackMinimaxVoice) featuring the same language boost and text normalization capabilities as the primary voice configuration.
+
+5. **Speaker Labeling**: Track multiple speakers in conversations with [`BotMessage.speakerLabel`](https://api.vapi.ai/api#:~:text=BotMessage.speakerLabel), providing stable speaker identification (e.g., "Speaker 1") for better conversation analysis and diarization.
+
+6. **Voice Region Support**: Choose optimal performance regions with Minimax's `worldwide` (default) or `china` regional settings for better latency and compliance with local regulations.
+
+<Info>
+    Language boost settings help the text-to-speech model better understand context and pronunciation for specific languages, resulting in more natural and accurate voice synthesis.
+</Info>
+
+## Voice Quality Features
+<CardGroup cols={2}>
+  <Card title="Multilingual Support" icon="globe">
+    Support for 40+ languages with automatic detection and language-specific optimizations for natural pronunciation.
+  </Card>
+  <Card title="Smart Text Processing" icon="type">
+    Intelligent normalization of numbers, dates, and formatted text for natural-sounding speech synthesis.
+  </Card>
+  <Card title="Performance Optimization" icon="zap">
+    Voice caching reduces latency for common phrases, while regional settings optimize for local performance.
+  </Card>
+  <Card title="Conversation Tracking" icon="users">
+    Speaker labeling and diarization support for multi-participant conversation analysis and management.
+  </Card>
+</CardGroup>

fern/changelog/2025-09-14.mdx

@@ -0,0 +1,35 @@
+# Squad Management & Session Enhancement
+
+1. **Squad-Based Sessions**: Organize your assistants into collaborative teams with [`Session.squad`](https://api.vapi.ai/api#:~:text=Session.squad) and [`Session.squadId`](https://api.vapi.ai/api#:~:text=Session.squadId). Sessions can now be associated with squads for team-based conversation management and coordinated assistant behavior.
+
+2. **Squad Chat Integration**: Enable squad-based chat conversations using [`Chat.squad`](https://api.vapi.ai/api#:~:text=Chat.squad) and [`Chat.squadId`](https://api.vapi.ai/api#:~:text=Chat.squadId). This allows multiple assistants to participate in or be aware of chat contexts for more sophisticated conversation handling.
+
+3. **Enhanced Session Creation**: Create squad-enabled sessions with [`CreateSessionDTO.squad`](https://api.vapi.ai/api#:~:text=CreateSessionDTO.squad) and [`CreateSessionDTO.squadId`](https://api.vapi.ai/api#:~:text=CreateSessionDTO.squadId), enabling persistent conversation contexts across multiple assistants and interaction types.
+
+4. **Chat Management by Squad**: Filter and organize chats by squad membership using [`GetChatPaginatedDTO.squadId`](https://api.vapi.ai/api#:~:text=GetChatPaginatedDTO.squadId) for better conversation management and team-based analytics.
+
+5. **Session Management by Squad**: Query sessions by squad association with [`GetSessionPaginatedDTO.squadId`](https://api.vapi.ai/api#:~:text=GetSessionPaginatedDTO.squadId), providing team-based session organization and management capabilities.
+
+6. **Full Message History**: Control conversation context retention with [`ArtifactPlan.fullMessageHistoryEnabled`](https://api.vapi.ai/api#:~:text=ArtifactPlan.fullMessageHistoryEnabled). When enabled, artifacts contain complete message history even after handoff context engineering, preserving full conversation flow for analysis.
+
+7. **Transfer Records**: Track warm transfer details with [`Artifact.transfers`](https://api.vapi.ai/api#:~:text=Artifact.transfers), providing comprehensive records of transfer destinations, transcripts, and status information for multi-assistant conversations.
+
+<Info>
+    Squad management enables sophisticated multi-assistant workflows where different specialists can handle different parts of a conversation while maintaining shared context and coordination.
+</Info>
+
+## Team Collaboration Features
+<CardGroup cols={2}>
+  <Card title="Multi-Assistant Coordination" icon="users">
+    Enable multiple assistants to work together within squads for specialized conversation handling and seamless handoffs.
+  </Card>
+  <Card title="Persistent Context" icon="database">
+    Maintain conversation context across squad members and session boundaries for continuous conversation experiences.
+  </Card>
+  <Card title="Team Analytics" icon="bar-chart">
+    Filter conversations, sessions, and analytics by squad membership for team-based performance insights and management.
+  </Card>
+  <Card title="Complete Audit Trails" icon="file-text">
+    Track all transfers and handoffs with detailed records including destinations, transcripts, and status information.
+  </Card>
+</CardGroup>