Eleven Labs Release Notes

Agents Platform

• Agent branching and deployments: Added a complete version control system for agents, enabling teams to create branches, iterate on agent configurations in isolation, and merge changes when ready. New endpoints include POST /v1/convai/agents/{agent_id}/branches for creating branches, POST /v1/convai/agents/{agent_id}/branches/{source_branch_id}/merge for merging, and POST /v1/convai/agents/{agent_id}/deployments for creating deployments. Drafts can also be created and deleted via the new drafts endpoints. See the Branches and Deployments documentation for details.
• WhatsApp account management: Added PATCH and DELETE endpoints for WhatsApp Business accounts, allowing you to update and remove connected WhatsApp accounts from agents.
• Conversation agent name: The Get conversation endpoint now returns agent_name in the response for easier identification of which agent handled a conversation.
• Error type tracking: Added error_type field to conversation event models for improved debugging and error categorization.

Knowledge Base …

Agents Platform

• Agent summaries endpoint: Added GET /v1/convai/agents/summaries endpoint for retrieving lightweight summaries of all agents in your workspace. This is useful for building agent selection interfaces without fetching full agent configurations.
• Delete batch calls: Added DELETE /v1/convai/batch-calling/{batch_id} endpoint for removing batch call jobs that are no longer needed.
• Branch filtering for conversations: Added optional branch_id query parameter to conversation endpoints including GET /v1/convai/conversation/get-signed-url, GET /v1/convai/conversation/token, and GET /v1/convai/conversations for filtering conversations by agent version branch.
• Spelling patience configuration: Added spelling_patience setting to turn configuration with values auto, low, medium, and high. Controls how long the agent waits before assuming the user has finished spelling something out character by character.
• Custom SIP headers for transfers: Added support for custom SIP headers when transferring calls via REFER. Configure custom_sip_headers arrays on phone number transfer configurations and workflow phone nodes for advanced telephony integrations.
• WhatsApp accounts on agents: Agent responses now include whatsapp_accounts array for viewing connected WhatsApp Business accounts. …

Scribe v2

We launched Scribe v2, the new state of the art transcription model. Learn more about Scribe v2 in the docs.

Agents Platform

• Timezone support for batch call scheduling: You can now select a timezone when scheduling batch calls. The scheduled time is converted to UTC based on your chosen timezone. Your browser's timezone is automatically selected by default, and validation prevents scheduling calls for times that have already passed in the selected timezone. This makes it easier to schedule outbound calls for the right time in your recipients' time zones.
• Knowledge Base source file URL: Added a new endpoint to retrieve the original source file URL for knowledge base documents, enabling direct access to uploaded files.
• LLM fallback cascade timeout: Added cascade_timeout_seconds configuration option for agent backup LLM configs, allowing control over how long to wait before cascading to the next LLM. Default is 8 seconds with an allowed range of 2-15 seconds.
• Soft timeout LLM-generated messages: Added use_llm_generated_message option to soft timeout configuration. When enabled, the agent will generate a contextual message using the LLM instead of using a predefined message when soft timeout triggers. …

Agents Platform

• Conversation ID in Twilio/SIP webhooks: The conversation_id is now included in webhook payloads when fetching conversation initiation client data for Twilio, SIP, and WhatsApp integrations. This enables real-time monitoring without polling the API to retrieve conversation IDs.
• Agent versioning fields: Agent response models now include version_id, branch_id, and main_branch_id fields (nullable strings) for tracking agent versions and branches.
• Conversation version tracking: Added version_id field to conversation models to identify the agent version used for each conversation.
• Batch call enhancements: The BatchCallRecipientStatus enum now includes dispatched status, and batch call responses include total_calls_finished field with improved defaults for tracking.

Voices

• Collection IDs: Voice models now include a collection_ids field (array of strings, nullable) indicating which collections a voice belongs to.
• Voice type filter expansion: The GET /v2/voices endpoint now supports saved as a voice_type filter value for retrieving non-default voices that have been added to a collection.

SDK Releases

Python SDK

• v2.28.0 - Updated API schema with agent versioning fields, voice collection IDs, batch call enhancements, and phone number labels

JavaScript SDK …

Agents Platform

• Real-time conversation monitoring: Enterprise users can now monitor agent conversations in real-time via WebSocket connection at /v1/convai/conversations/{id}/monitor. Features include cached conversation history, selective event streaming, and control commands (end call, barge-in, human takeover toggle, send messages).
• Hinglish language support: Added hinglish_mode configuration option for agents. When enabled and the agent's language is Hindi, responses will be in Hinglish (Hindi-English mix).
• Dynamic variables in voicemail messages: Voicemail detection messages now support dynamic variable substitution, enabling personalized messages when voicemail is detected.
• Localized widget terms and conditions: Chat widget terms and conditions can now be localized per language, configured through language presets.
• Widget conversation mode toggle: Added conversation_mode_toggle_enabled option to widget configuration for controlling conversation mode switching.
• Batch call retry tracking: Added retry_count field to batch call status responses for tracking retry attempts.

Text to Speech

• Quality preset credit costs: Quality preset descriptions now include credit cost information: high (+20%), ultra (+50%), and ultra_lossless (+100%).

WhatsApp Integration …

Agents Platform

• Twilio call registration: Added new endpoint to register Twilio calls and receive TwiML response for agent conversations.

Music Generation

• Word timestamps: Added word-level timestamps to music generation with detailed responses. Enable with the with_timestamps parameter.
• Fine-tuning support: Added finetune_id parameter to music generation endpoints for using custom music fine-tunes.
• C2PA content signing: Added sign_with_c2pa option to music generation requests for content authenticity and provenance.

Voices

• Voice settings expansion: Added voice_stability, voice_similarity, and voice_style parameters to voice endpoints for more granular control.
• Pagination improvements: Clarified pagination behavior for Get all voices v2 endpoint with better documentation of next_page_token, has_more, and total_count fields.

ElevenCreative Studio

• Voice settings override: Added voice_settings parameter to project creation for per-project voice configuration.

Webhook Tools

• Custom request headers: Added request_headers support to Webhook tool configuration for custom authentication and metadata.

Billing

• Schema improvements: Introduced BillingPeriod and CharacterRefreshPeriod enums for better type safety in subscription models. …

Agents Platform

• New LLM support: Extended LLM options with gemini-3-pro-preview model.
• Out-of-band DTMF: Added use_out_of_band_dtmf flag to PlayDTMFToolConfig for improved DTMF tone handling in telephony integrations.
• Validation improvements: Reduced maximum items for agent test submission from 200 to 100 for both ResubmitTestsRequestModel.test_run_ids and RunAgentTestsRequestModel.tests to improve performance and reliability.
• Live analytics: Added new analytics endpoint to retrieve real-time count of active conversations.
• Custom LLM API types: Added api_type field to CustomLLM schema with enum CustomLLMAPIType supporting chat_completions and responses options for flexible LLM integration.
• Conversation filtering: Enhanced conversation filtering with support for feedback ratings and comments.

Dubbing

• Speaker management: Added ability to create new speakers for dubbing resources and migrate segments between speakers.
• Speaker updates: Added speaker_name field support when updating dubbing speakers.
• Transcript defaults: Relaxed transcript schema requirements with defaults for text, word_type, and speaker_id fields, and removed some required field constraints for easier integration.
• Language code support: Clarified that source_lang and target_lang fields expect ISO 639-1/3 codes.

ElevenCreative Studio …

Dubbing

• JSON transcript format: Added support for JSON format transcripts in the GET /v1/dubbing/{dubbing_id}/transcript/{language_code} endpoint, returning detailed word-level and character-level timing data via DubbingTranscriptResponseModel.

Agents Platform

• GPT-5.1 support: Added support for gpt-5.1 and gpt-5.1-2025-11-13 LLM models in agent configurations.
• Reasoning effort control: Added none option to LLMReasoningEffort for improved control over model reasoning behavior.
• Agent tool request events: Added agent_tool_request to ClientEvent for better tool interaction tracking in real-time conversations.
• Out-of-band DTMF: Added use_out_of_band_dtmf flag to PlayDTMFToolConfig for improved telephony integration.
• LLM model additions: Extended LLM enum with gemini-3-pro-preview for enhanced conversational AI capabilities.

Speech to Text

• Scribe quota exceeded events: Added quota_exceeded event in SDKs and include_timestamps parameter for better quota management and timestamp control.

ElevenCreative Studio

• Chapter voice tracking: Added voice_ids field to chapter response models for better voice management in multi-voice projects.
• Project image assets: Introduced ProjectImageResponseModel for image asset support, now included in ProjectResponseModel.assets. …

Scribe v2 Realtime

Scribe v2 Realtime, our fastest and most accurate live speech recognition model, has officially launched. It delivers state-of-the-art accuracy in over 92 languages with an ultra-low 150ms of latency. Read more here.

Agents Platform

• Widget end-call feedback: Widgets now support customizable end-of-call feedback collection with optional rating and comment fields, enabling you to gather user feedback directly within your agent conversations.
• Soft timeout configuration: Added comprehensive soft timeout settings to control how agents handle pauses in conversation, with configurable prompts and behavior across turn and conversation levels.
• Advanced conversation filtering: The conversations endpoint now supports filtering by call duration range, evaluation parameters, data collection parameters, and specific tool names.
• Enhanced conversation feedback: Updated conversation feedback system with structured feedback types including ratings and comments for better user experience tracking.

ElevenCreative Studio

• Extended project metadata: Projects now include comprehensive asset tracking with video thumbnails, external audio references, and enhanced snapshot information including audio duration.

Billing …

UI

• Improved notifications: Updated notification system to display notifications relevant to the active platform you're viewing, providing a more focused and contextual experience.

Agents Platform

• Dynamic variable transfer destinations: Agent transfers now support dynamic variables for phone numbers and SIP URIs, enabling runtime-determined transfer destinations based on conversation context.
• MCP tool configuration overrides: Added the ability to create, update, retrieve, and delete custom configuration overrides for specific MCP Server tools, allowing fine-grained control over tool behavior and parameters.

Text to Dialogue

• Timestamps and voice segments: Text to Dialogue now supports timestamped outputs with character-level alignment and voice segment tracking, making it easier to synchronize dialogue with animations or subtitles.

Music API

• Stem separation: Added new stem separation endpoint to isolate different audio components (vocals, drums, bass, instruments) from existing music tracks.
• Increased prompt length: Music generation now supports prompts up to 4,100 characters, with individual lyric lines supporting up to 200 characters.

Security

• Single-use tokens: Introduced time-limited single-use token generation for secure operations, providing enhanced security for sensitive API operations.

SDK Releases

JavaScript SDK …

Workspaces

• External voice sharing controls: External voice sharing and Voice Library publishing can now be disabled via workspace groups. Internal (in-organization) voice sharing is not impacted.

Agents Platform

• Multi-environment agent management: The Agents CLI now supports the --env flag, allowing you to manage agents deployed to multiple accounts simultaneously.
• Search functionality: Added search parameter to the Get conversations endpoint for filtering conversations.

SDK Releases

JavaScript SDK

• v2.20.1 - Fixed Music API type definitions and added Realtime Scribe language code support
• v2.20.0 - Added helper methods for Realtime Scribe with improved ASR capabilities. This is a beta feature and not widely available yet

Python SDK

• v2.20.1 - Fixed file parameter handling in client methods
• v2.20.0 - API schema updates and improvements
• v2.19.0 - Added helper method for Scribe Realtime with improved ASR capabilities. This is a beta feature and not widely available yet

Agents CLI …

Agents Platform

• Genesys configuration overrides: Genesys users can now override agent configuration directly from their Genesys flows using session variables. Use system__override_system_prompt, system__override_first_message, and system__override_language to customize the agent's behavior, first message, and language for specific call flows without modifying the base agent configuration.

• Batch call ID copying: Added copy-to-clipboard functionality for batch call IDs throughout the Agents Platform interface. Click on batch call IDs in the batch call details page or conversation history to quickly copy them for API interactions, support requests, or debugging.

SDK Releases

Python SDK

• v2.18.0 - Added streaming support for Agents Platform conversations via the Conversation class, Music API inpainting with new output format options, and Agent Workflows support including workflow models, node types, and conditional logic operators (AND, OR, equals, greater than, less than, etc.)

JavaScript SDK …

Agents Platform

• LLM overrides: Added support for overriding an agent's LLM during a conversation, enabling you to specify a different language model on a per-conversation basis. This is useful for testing different models or accommodating specific requirements while maintaining HIPAA and data residency compliance.
• Post-call webhook failures: Added the option to send post-call webhook events in the event of a phone call failure. This allows you to track and respond to failed call attempts through your webhook endpoint, providing better visibility into call issues.

SDK Releases

Python SDK

• v2.18.0 - Added support for streaming, Music API inpainting, and Agent Workflows

JavaScript SDK

• v2.19.0 - Added support for Music API inpainting and Agent Workflows
• v2.18.0 - API schema updates

Client Packages

• @elevenlabs/agents-cli@0.5.0 - Significantly reworked agents pull command with bugfixes and improvements
• @elevenlabs/react@0.8.0 - Fixed import issues …

Agents Platform

• Gemini 2.5 Flash Preview models: Added support for gemini-2.5-flash-preview-09-2025 and gemini-2.5-flash-lite-preview-09-2025 LLM models, providing access to the latest September 2025 preview versions of Google's Gemini 2.5 Flash models.
• Claude Sonnet 4.5: Added support for claude-sonnet-4-5 and claude-sonnet-4-5@20250929 models, enabling access to the latest Claude Sonnet 4.5 model released on September 29, 2025.
• Test invocations listing: Added new GET /v1/convai/test-invocations endpoint to list all test invocations with pagination support. Includes agent_id filter, page_size parameter (default 30, max 100), and cursor for pagination. Response includes test run counts, pass/fail statistics, and titles.
• Agent archiving: Added archived field (boolean, default false) to agent platform settings, allowing agents to be archived without deletion while keeping them out of active agent lists.
• MCP Server interruption control: Added disable_interruptions field (boolean, default false) to MCP server configuration, preventing user interruptions during tool execution for more reliable tool completion.
• Streaming agent responses: Added agent_chat_response_part WebSocket event type for receiving partial agent chat responses in real-time during streaming conversations. …

v1 TTS model deprecation

The eleven_monolingual_v1 and eleven_multilingual_v1 models are deprecated and will be removed on December 15th, 2025. Please migrate to newer models for continued service.

Agents Platform

• Workflow Expressions: Workflows now support complex expressions that allow for defining deterministic conditions using logical operators, dynamic variables and LLM evaluation. This enables more sophisticated agent logic and decision-making capabilities.
• MCP Server Interrupt Control: Added option to disable interruptions during all tool calls for MCP Servers, providing better control over agent behavior during tool execution.
• Audio Alignment Data: Agents now have a flag to enable alignment data in audio events, useful for audio-text synchronization use cases such as lip sync applications.
• Ignore Default Personality Setting: The Agents Platform configuration page now includes a checkbox to toggle whether agents should ignore the default helpful personality, giving developers more control over agent behavior.

Speech to Text

• Fixed Base64 Encoding Flag: Resolved an issue where the is_base64_encoded flag in STT responses was incorrectly set to false for PDF and DOCX formats, even when content was actually base64 encoded.

SDK Releases

JavaScript SDK

• v2.16.0: Updated with latest API schema changes from September 19, 2025. …

Productions launch

Introducing Productions - our new managed service offering for ordering human-edited content that looks, sounds and feels natural. Made for creators and media businesses.

Our network of linguists and audio professionals offer end-to-end production quality for:

• Dubbing
• Captions and subtitles
• Transcription
• Audiobooks

You can order a project directly from the 'Productions' page in your ElevenLabs account, or by emailing productions@elevenlabs.io. Pricing starts at $2/minute, contact us for more details.

Agents Platform

• MCP pre-tool speech: Added support for configuring tools extracted from an MCP Server to require pre-tool execution speech. This enhancement allows agents to provide verbal context before executing specific tools, improving the conversational flow during tool usage.

• ElevenLabs hosted LLMs: Added support for ElevenLabs hosted LLMs which unlock lower latency by running on ElevenLabs infrastructure alongside Speech to Text and Text to Speech services.

• Enum values for tool parameters: Added support for specifying a tool's parameters as enum values for greater control

SDK Releases

JavaScript SDK …

Text to Speech

• WebSocket output format: Added support for specifying output format in the first message of a WebSocket connection, providing greater flexibility for real-time audio streaming workflows.

Agents Platform

• First message interruption control: Added disable_first_message_interruptions setting to prevent agents from being interrupted during important opening messages like legal disclaimers.

MCP Server

• Version 0.8.1: Added data residency support.

SDK Releases

JavaScript SDK

• v2.15.0 - Added new Text to Voice Remix endpoint

Python SDK

• v2.15.1 - Fixed conversation authentication issue and added asyncio event loop support
• v2.15.0 - Added new Text to Voice Remix endpoint and fixed Pydantic issues

Packages

• @elevenlabs/react@0.6.2 - Added correction and MCP tool call events
• @elevenlabs/client@0.6.2 - Added correction and MCP tool call events
• @elevenlabs/react-native@0.3.1 - Added correction and MCP tool call events

API

Text to Speech

• Language code support: All Text to Speech models now support language codes for improved output. Normalization has been enabled for Eleven v3, Flash, and Turbo models to enhance audio quality and consistency.

Agents Platform

• Multi-voice agent history: Messages from multi-voice agents are now displayed in conversation history with clear separation by voice, making it easier to follow which voice spoke which part of a conversation.

SDK Releases

JavaScript SDK

• v2.15.0 - Adds support for new voice remix functionality

Python SDK

• v2.15.0 - Adds support for new voice remix functionality. Also fixed an issue with Pydantic.

React Components

• @elevenlabs/react@0.6.1 - Fix output bytes and device input/output switching
• @elevenlabs/client@0.6.1 - Fix output bytes and device input/output switching

MCP Server

• v0.7.0 - Latest release of the ElevenLabs MCP Server with new features and improvements for Claude Desktop integration. Includes new loop parameter for SFX generation.

API

Agents Platform

• Gemini 2.5 Flash Lite HIPAA compliance: Added Gemini 2.5 Flash Lite to the list of HIPAA approved models for compliant conversations when a BAA is signed and zero-retention mode is enabled.
• Conversation ID in signed URLs: Added support for including conversation IDs in signed URL requests, providing better tracking and identification capabilities for conversation audio access.

SDK Releases

JavaScript SDK

• v2.13.0 - Released August 29, 2025. Adds support for new loop parameter in SFX.

Python SDK

• v2.13.0 - Released August 29, 2025. Adds support for new loop parameter in SFX.

ConvAI packages

• @elevenlabs/react v0.6.0 and @elevenlabs/client v0.6.0 - Released August 29, 2025. Fixed setVolume functionality, added client tool debugging, and added audio device controls.

MCP Server

• ElevenLabs MCP Server v0.6.0 - Released August 26, 2025. Fixed diarization functionality in speech-to-text and added music generation endpoints.

API

Agents Platform

• Agent testing framework: Introduced a comprehensive testing framework for ElevenLabs agents, allowing developers to create, manage, and execute automated tests for their agents. This includes test creation, execution tracking, and result analysis capabilities.
• Test invocation management: Added support for resubmitting failed test invocations and viewing detailed test results to help developers debug and improve their agents.
• Enhanced agent configuration: Improved agent creation and management with additional workspace override capabilities and refined platform settings.

Text to Speech

• Pronunciation dictionary updates: Added support for updating pronunciation dictionaries with PATCH operations, enabling more flexible dictionary management.
• Enhanced timestamp support: Improved timestamp generation for text-to-speech conversions with better alignment data and streaming capabilities.

SDK Releases

• TypeScript SDK v2.12.2: Updated with the latest API schema changes, including full support for the new agent testing endpoints and enhanced Agents Platform capabilities.
• Python SDK v2.12.1: Released with complete support for all new API features, including agent testing framework and improved workspace resource management.

API