September 14, 2026
ElevenAgents
Section titled “ElevenAgents”- Call queueing and hold audio: Added call queueing for agents at their concurrency limit. Queued callers hear hold audio and receive
queue_statusevents until they are admitted or time out. New API methods upload or delete custom hold audio. - Test version metadata: Agent test results now identify the agent version they ran against and whether they included draft changes. Invocation summaries also indicate when partial resubmissions ran against different versions.
- Model and account metadata: Agent LLM configuration now accepts
gemini-3.8-flash. WhatsApp account responses identify whether the account uses the Cloud API or coexistence signup flow.
- Music v2.5 API support: Music endpoints and SDKs now accept
music_v2_5. Composition plan generation chunks support up to 6,132 characters, with up to 30 lines of 200 characters each.
SDK releases
Section titled “SDK releases”JavaScript SDK
Section titled “JavaScript SDK”- v2.68.0 - Added methods to upload and delete agent hold audio. Regenerated types for call queueing, procedure references, draft branch creation, RAG query limits, test version metadata,
gemini-3.8-flashandmusic_v2_5.
Python SDK
Section titled “Python SDK”- v2.68.0 - Added synchronous and asynchronous methods to upload and delete agent hold audio. Regenerated types for call queueing, procedures, draft branch creation, RAG query limits, test version metadata,
gemini-3.8-flashandmusic_v2_5. Thecompose_detailedwrapper now accepts every generated music model ID and both prompt- and chunk-based composition plans.
iOS SDK
Section titled “iOS SDK”- v3.3.1 - Reconciles agent messages by
response_idinstead ofevent_id, preserving multiple responses around tool calls and preventing streamed parts or corrections from attaching to the wrong message.
ElevenLabs CLI
Section titled “ElevenLabs CLI”- v1.3.0 - Added optional
--intentmetadata to generated API commands. Addedelevenlabs feedback missing-capabilityto report workflows that the CLI does not support.
Packages
Section titled “Packages”- @elevenlabs/convai-widget-core@0.18.2 and @elevenlabs/convai-widget-embed@0.18.2 - Fixed duplicate agent replies after tool and procedure calls. Streamed text replies retain Markdown formatting, and voice conversations no longer render unspoken streamed chat parts. The embed package also limits
user_activityevents to once per second and clears the limit when a session ends.
View API changes
New endpoints
ElevenAgents
- Upload agent hold audio -
POST /v1/convai/agents/{agent_id}/hold-audio- Accepts required
hold_audio_file(binary) as multipart form data. The file must be MP3 or WAV, up to 40 MB and 180 seconds. - Returns required
agent_id(string) andhold_audio(AgentHoldAudioConfig). The hold-audio object requiresaudio_path,audio_url,original_filename(strings),duration_secs(number) andsize_bytes(integer).
- Accepts required
- Delete agent hold audio -
DELETE /v1/convai/agents/{agent_id}/hold-audio- Removes the custom clip so queued callers hear the default hold tone.
- Returns required
agent_id(string).
Updated endpoints
ElevenAgents
- Create agent, Get agent and Update agent
- Added optional
platform_settings.queueing_config(AgentQueueingConfig) withenabled(boolean, defaultfalse) andwait_timeout_seconds(integer, 1–1,800, default180). - Queue configuration responses include read-only nullable
hold_audio(AgentHoldAudioConfig). Agent responses also include optionaldefault_hold_audio_url(string).
- Added optional
- Get conversation -
GET /v1/convai/conversations/{conversation_id}- Metadata adds optional nullable
queue_wait_secs(number). Queue time is excluded fromcall_duration_secsand billed time.
- Metadata adds optional nullable
- Create agent branch -
POST /v1/convai/agents/{agent_id}/branches- Added optional
include_draft(boolean, defaultfalse). Whentrue, a branch created from the branch tip includes the caller's draft procedure set.
- Added optional
- List procedures and procedure create, get and draft endpoints
- List items add optional
referenced_tool_ids,referenced_kb_ids,referenced_procedure_idsandreferenced_dynamic_variables(string arrays). - Agent and merge-preview responses add optional
procedures, a map keyed by procedure ID.
- List items add optional
- List test invocations -
GET /v1/convai/test-invocations- Added optional nullable
search(string) to filter tests and folders by name.
- Added optional nullable
- Run tests on agent and Get test invocation
- Test suite invocation and unit test run responses add optional nullable
version_id(string) and optionalran_against_draft(boolean, defaultfalse). - Invocation summaries also add optional
runs_diverged_from_version(boolean, defaultfalse).
- Test suite invocation and unit test run responses add optional nullable
POST /v1/convai/agents/{agent_id}/knowledge-base/rag-query- The request body adds optional nullable
max_documents_length(integer, 1–50,000) andmax_retrieved_rag_chunks_count(integer, 1–20) query-specific overrides.
- The request body adds optional nullable
- Get WhatsApp account -
GET /v1/convai/whatsapp-accounts/{phone_number_id}- Added optional
account_type(WhatsAppAccountType, defaultcloud_api). Values arecloud_apiandcoexistence.
- Added optional
- Calculate expected LLM usage -
POST /v1/convai/agent/{agent_id}/llm-usage/calculateLLMUsageCalculatorLLMResponseModeladds requiredprice_per_message(number).
Music
MusicModelIDaddsmusic_v2_5forPOST /v1/music/video-to-music,POST /v1/music/plan, Compose music,POST /v1/music/detailed,POST /v1/music/detailed/stream,POST /v1/music/stream,POST /v1/music/uploadandPOST /v1/music/finetunes.GenerationChunk.textincreasesmaxLengthfrom 6,000 to 6,132 characters forPOST /v1/music/plan,POST /v1/music,POST /v1/music/detailed,POST /v1/music/detailed/stream,POST /v1/music/streamandPOST /v1/music/upload. Section names can contain 1–100 characters, and chunks can contain up to 30 lines with 200 characters per line.
Schema changes
ElevenAgents
- Real-time server events add
queue_status. Its requiredqueue_status_event.statusacceptswaiting,admittedortimed_out;timed_outprecedes WebSocket close code4300. Hold audio continues to arrive as regularaudioevents. - The
LLMenum addsgemini-3.8-flash, including conversation initiation overrides. - Conversation system tool result unions add
start_procedure_success,start_procedure_error,end_procedure_successandend_procedure_error. FeaturesUsageCommonModeladds optionalfreeform_procedureandstructured_procedurefeature statuses.