Update LiveKit dependencies (non-major) to v1.4.1

[renovate]

ℹ️ Note

This PR body was truncated due to platform limits.

This PR contains the following updates:

Package	Change	Age	Confidence
livekit-agents	==1.2.7 → ==1.4.1
livekit-plugins-deepgram	==1.2.7 → ==1.4.1
livekit-plugins-openai	==1.2.7 → ==1.4.1
livekit-plugins-silero	==1.2.7 → ==1.4.1

---

Release Notes

livekit/agents (livekit-agents)

v1.4.0

Compare Source

Python 3.14 Support & Python 3.9 Dropped

This release adds Python 3.14 support and drops Python 3.9. The minimum supported version is now Python 3.10.

Tool Improvements

Tools and toolsets now have stable unique IDs, making it possible to reference and filter tools programmatically. Changes to agent configuration (instructions, tools) are now tracked in conversation history via AgentConfigUpdate.

LLMStream.collect() API

A new LLMStream.collect() API makes it significantly easier to use LLMs outside of AgentSession. You can now call an LLM, collect the full response, and execute tool calls with a straightforward API — useful for background tasks, pre-processing, or any workflow where you need LLM capabilities without the full voice agent pipeline.

from livekit.agents import llm

response = await my_llm.chat(chat_ctx=ctx, tools=tools).collect()

for tc in response.tool_calls:
    result = await llm.execute_function_call(tc, tool_ctx)
    ctx.insert(result.fnc_call)
    if result.fnc_call_out:
        ctx.insert(result.fnc_call_out)

Manual Turn Detection for Realtime Models

Realtime models now support commit_user_turn, enabling turn_detection="manual" mode. This gives you full control over when user turns are committed — useful for push-to-talk interfaces or scenarios where automatic VAD-based turn detection isn't ideal.

@​ctx.room.local_participant.register_rpc_method("end_turn")
async def end_turn(data: rtc.RpcInvocationData):
    session.input.set_audio_enabled(False)
    session.commit_user_turn(
        transcript_timeout=10.0,
        stt_flush_duration=2.0,
    )

Job Migration on Reconnection

When the agent server temporarily loses connection and reconnects, active jobs are now automatically migrated rather than being dropped. This significantly improves reliability during transient network issues.

False Interruption Fix

Fixed a bug where late end-of-speech events could trigger duplicate false interruption timers, causing the agent to incorrectly stop speaking. The agent now properly deduplicates these events and tracks STT completion state more reliably.

New Providers & Plugins

• xAI Responses LLM — Use xAI's Responses API via xai.responses.LLM()
• Azure OpenAI Responses — Azure-hosted Responses API via azure.responses.LLM(), with support for deployments and Azure auth
• Camb.ai TTS — New TTS plugin powered by the MARS model family (mars-flash, mars-pro, mars-instruct), with voice selection, language control, and style instructions
• Avatario Avatar — Virtual avatar plugin with session management and API client

What's Changed

• feat(azure/stt): TrueText post processing option added to STTOptions by @​rafallezanko in #​4557
• chore(README): remove STT and LLM API key configuration from LemonSlice example as not needed. by @​codeSTACKr in #​4589
• fix: Add thread-safe initialization to _DefaultLoadCalc singleton by @​darshankparmar in #​4585
• add missing plugins to dependencies by @​tinalenguyen in #​4593
• _setup_cloud_tracer still overrides TracerProviders due to checking the wrong base class by @​hudson-worden in #​4584
• fix(google): add thought_signature support for Gemini 2.5 models by @​gdoermann in #​4595
• remove shortcut inference STT model name by @​longcw in #​4594
• Increase read_bufsize in minimax tts plugin by @​jose-speak in #​4590
• refactor(rtzr): FlushSentinel-based segment control and type safety improvements by @​kimdwkimdw in #​4565
• improve EndCallTool by @​longcw in #​4563
• Fix: Add 'required' field to function_tool schema for Groq compatibility by @​VinayJogani14 in #​4613
• fix: avoid modifying original raw tool description by @​davidzhao in #​4616
• continue instead of return in InferenceProcExecutor loop by @​chenghao-mou in #​4612
• add xai responses llm by @​tinalenguyen in #​4618
• move xAI tools to separate file by @​tinalenguyen in #​4624
• (xAI): backward compatibility for tools by @​tinalenguyen in #​4625
• update inference models to match the latest by @​davidzhao in #​4597
• AssemblyAI added EU streaming endpoint option by @​ftsef in #​4571
• feat: Add Camb.ai TTS plugin by @​eRuaro in #​4442
• prevent duplicate false interruption due to late end of speech by @​chenghao-mou in #​4621
• feat: add customization bithuman gpu avatar endpoint handling by @​CathyL0 in #​4390
• plugin/liveavatar implement sandbox on liveavatar by @​arthurnumen in #​4635
• add asyncai to pyproject by @​tinalenguyen in #​4636
• feat: avatario avatar plugin by @​Saksham209 in #​4114
• add azure openai responses by @​tinalenguyen in #​4619
• (openai realtime): add truncation param by @​tinalenguyen in #​4642
• fix: 11Labs Scribe v2 model not working with EOT prediction model by @​Ludobaka in #​4601
• (taskgroup): support on_complete callback functions by @​tinalenguyen in #​4628
• add id to tools by @​theomonnom in #​4653
• allow 499 retry by @​chenghao-mou in #​4637
• AGT-2474: add commit user turn support for realtime models by @​chenghao-mou in #​4622
• fix(liveavatar): emit playback_finished on AudioSegmentEnd by @​MSameerAbbas in #​4669
• add AgentConfigUpdate & initial judges by @​theomonnom in #​4547
• fix tests & ruff by @​theomonnom in #​4672
• (minimax): add language boost param by @​tinalenguyen in #​4667
• remove accidentally committed files by @​theomonnom in #​4673
• fix duplicated openai realtime remote content by @​longcw in #​4657
• use text streams & custom rpc logic by @​theomonnom in #​4677
• remove chat_ctx size limit by @​theomonnom in #​4678
• clean up metrics export from traces by @​davidzhao in #​4679
• LLMStream.collect API & external easier tool executions by @​theomonnom in #​4680
• update openai responses default model by @​tinalenguyen in #​4681
• fix(google): improve error message for model/API mismatch in Realtime API by @​cdutr in #​4611
• fix keyterm in Deepgram by @​chenghao-mou in #​4684
• Expose ws close code and error messages by @​chenghao-mou in #​4683
• fix: improve handling of 499 status code by @​davidzhao in #​4685
• support wrapped tools with a warning message by @​longcw in #​4674
• fix(transcription): prevent stale synchronizer impls (#​4486) by @​furious-luke in #​4686
• Add rtzr plugin to optional dependencies by @​zach-iee in #​4631
• feat(langgraph): add custom stream mode support in LangChain LLMAdapter by @​keenranger in #​4511
• Add room deletion timeout and cancellation by @​chenghao-mou in #​4638
• add TaskCompletedEvent import by @​tinalenguyen in #​4688
• prevent tool cancellation when AgentTask is called inside it by @​longcw in #​4586
• fix gemini live tool execution interrupted by generation_complete event by @​longcw in #​4699
• add STT usage for google by @​chenghao-mou in #​4599
• fix: commit user turn with STT and realtime by @​chenghao-mou in #​4663
• add exclude_config_update to ChatContext copy by @​longcw in #​4700
• add require_confirmation param for built-in tasks by @​tinalenguyen in #​4698
• Fix wrong "timestamp" parameter in livekit-plugins-spitch stt.py by @​pabloFuente in #​4702
• Update readme and examples to use deepgram nova-3 by @​bcherry in #​4697
• set exclude_config_update by @​longcw in #​4709
• Restore Python 3.14 support by updating livekit-blingfire to 1.1 by @​Abivarman123 in #​4710
• add ChatContext.messages() by @​theomonnom in #​4712
• migrate jobs on reconnection by @​theomonnom in #​4711
• use ChatMessage.messages() where applicable by @​theomonnom in #​4713
• chore(docs): ditch the v0 docs and promote v1 docs to main path by @​rektdeckard in #​4695
• filter tools by id by @​tinalenguyen in #​4723
• support python 3.14 by @​theomonnom in #​4727
• Fix: Added stt lang parsing and tts voice parsing to the constructors by @​adrian-cowham in #​4726
• fix (liveavatar): restore interruption handling and track avatar speaking state by @​tinalenguyen in #​4725
• update padding warning message and silence subsequent ones by @​chenghao-mou in #​4733
• fix: Add default google tts model selection for backward compatibility by @​chenghao-mou in #​4731
• fix uv lock file & drop python 3.9 support & upgrade dependencies by @​theomonnom in #​4728
• automatically close openai client by @​theomonnom in #​4735
• update gitignore by @​theomonnom in #​4737
• fix speaking_rate inference by @​theomonnom in #​4738
• livekit-agents 1.4.0 by @​theomonnom in #​4739

New Contributors

• @​codeSTACKr made their first contribution in #​4589
• @​hudson-worden made their first contribution in #​4584
• @​gdoermann made their first contribution in #​4595
• @​jose-speak made their first contribution in #​4590
• @​VinayJogani14 made their first contribution in #​4613
• @​ftsef made their first contribution in #​4571
• @​eRuaro made their first contribution in #​4442
• @​arthurnumen made their first contribution in #​4635
• @​Saksham209 made their first contribution in #​4114
• @​Ludobaka made their first contribution in #​4601
• @​cdutr made their first contribution in #​4611
• @​keenranger made their first contribution in #​4511
• @​Abivarman123 made their first contribution in #​4710

Full Changelog: https://github.com/livekit/agents/compare/livekit-agents@1.3.12...livekit-agents@1.4.0

v1.3.12

Compare Source

What's Changed

• improve text mode CLI rendering by @​theomonnom in #​4522
• fix Worker.aclose raising RuntimeError by @​theomonnom in #​4523
• better cli rendering for audio by @​theomonnom in #​4524
• fix frame capture order and add playback start callback in console mode by @​chenghao-mou in #​4516
• Add Connector to default participant kinds by @​cnderrauber in #​4526
• add support for language detection for assembly ai by @​chenghao-mou in #​4527
• Support static context in integration with langchain by @​benlangfeld in #​4504
• feat(google): add warnings when system messages are dropped in Gemini realtime model by @​dhruvnigam93 in #​4513
• chore: change deprecated cartesia voice id by @​davidzhao in #​4528
• #​4481 Added Opus and PCM encoding to ElevenLabs TTS by @​rafallezanko in #​4525
• interrupt the same speech handle by @​chenghao-mou in #​4536
• pin livekit-rtc version by @​theomonnom in #​4531
• fix(elevenlabs/stt): allow specifying scribe_v2 non-realtime model by @​bml1g12 in #​4515
• add reasoning param for openai responses LLM by @​tinalenguyen in #​4548
• Defensive fixes by @​chenghao-mou in #​4546
• LemonSlice Plugin by @​jp-lemon in #​4539
• feat (google STT): support profanity filter by @​tinalenguyen in #​4573
• fix(baseten): correct metadata and response field names for STT by @​toubatbrian in #​4572
• drop frames when the ConsoleAudioInput is detached by @​longcw in #​4576
• fix audio recording in console mode by @​longcw in #​4575
• Chatterbox model support by @​plangary in #​4541
• Inworld websocket improvements by @​cshape in #​4533
• fix(deepgram): expose close code and reason on unexpected disconnects by @​vadimatmurphy in #​4569
• playback started call for DataStreamAudioOutput and QueueAudioOutput by @​chenghao-mou in #​4570
• feat(azure): add lexicon_uri option to TTS by @​zach-iee in #​4485
• feat(tts): integrate AsyncAI TTS engine into livekit by @​ashotbagh in #​3596
• Simplismart Integration in Livekit by @​Tushar-ml in #​4349
• handle invalid bytes error by @​chenghao-mou in #​4579
• Fixes #​4388: Correct transcription_delay metric calculation in STT turn detec… by @​devbyteai in #​4396
• fix(mcp): Error message based on text attribute instead of str(part) by @​rafallezanko in #​4582
• livekit-agents 1.3.12 by @​theomonnom in #​4583

New Contributors

• @​cnderrauber made their first contribution in #​4526
• @​benlangfeld made their first contribution in #​4504
• @​dhruvnigam93 made their first contribution in #​4513
• @​jp-lemon made their first contribution in #​4539
• @​vadimatmurphy made their first contribution in #​4569
• @​zach-iee made their first contribution in #​4485
• @​ashotbagh made their first contribution in #​3596
• @​Tushar-ml made their first contribution in #​4349
• @​devbyteai made their first contribution in #​4396

Full Changelog: https://github.com/livekit/agents/compare/livekit-agents@1.3.11...livekit-agents@1.3.12

v1.3.11

Compare Source

What's Changed

• Add allowed_tools and transport_type parameters to MCPServerHTTP by @​wasaybaig201 in #​4365
• better transport_type type in MCPServer by @​theomonnom in #​4375
• fix typo: double the - in multiple livekit-plugin providers by @​ChristianBernhard in #​4358
• feat(AWS STT): use ChainedIdentityResolver by @​itskyf in #​4356
• support next_in_chain for RoomIO text output by @​longcw in #​4353
• standardize Tool interface by @​longcw in #​4368
• add tests for function tool parsing and execution by @​longcw in #​4384
• add EndCallTool by @​longcw in #​4377
• chore(google): update doc string to reflect default realtime models by @​davidzhao in #​4398
• fix AttributeError in NVIDIA Riva STT by @​gau-nernst in #​4391
• fix aws credentials when using env vals by @​chenghao-mou in #​4403
• fix (gemini streaming tts): change default audio encoding and voice by @​tinalenguyen in #​4393
• fix (mistral-ai): add flexibility for timestamps by @​tinalenguyen in #​4404
• remove shutdown models from google gemini live by @​tinalenguyen in #​4421
• update groq tts models, voices, and defaults by @​tinalenguyen in #​4422
• fix (google stt): set enable_word_time_offsets to False for chirp 3 by @​tinalenguyen in #​4420
• fix: return 503 health check when worker fails to connect to LiveKit by @​rusg77 in #​4419
• Add retrieval config support for google LLM by @​chenghao-mou in #​4408
• allow pushing frames to VAD when agent speech is uninterruptible by @​chenghao-mou in #​4418
• Add extra comments about Google model deprecation by @​chenghao-mou in #​4424
• fix (gemini filesearch): require only filestore names by @​tinalenguyen in #​4428
• update docker dependencies by @​chenghao-mou in #​4431
• chore: minor fixup of console room name by @​davidzhao in #​4433
• Inference: Improved support for mid session TTS updates by @​adrian-cowham in #​4412
• fix: acquire lock in _DefaultLoadCalc.get_load() to prevent race condition by @​martin-purplefish in #​4435
• fix vad rnn state by @​theomonnom in #​4437
• restore old behavior by @​chenghao-mou in #​4434
• fix: avoid double RoomIO.aclose during shutdown by @​darshankparmar in #​4446
• add connect CLI command by @​tinalenguyen in #​4452
• fix function call created_at by @​longcw in #​4453
• Update STT tests and add batch recognition flag by @​chenghao-mou in #​4425
• allow tests on external PRs when triggered by members by @​chenghao-mou in #​4456
• Adding model query param to the STT and TTS websocket connection string. by @​adrian-cowham in #​4457
• refactor connect CLI command by @​tinalenguyen in #​4458
• OpenAI Responses API Plugin by @​tinalenguyen in #​4192
• fix transcription truncate when agent is interrupted in console mode by @​longcw in #​4473
• feat(rtzr): add keyword boosting to streaming STT by @​lalq in #​4405
• Revise AWS Plugin README for accuracy and clarity by @​guiruggiero in #​4468
• close log_handler when process initialize failed by @​longcw in #​4472
• feat(deepgram): make vad_events configurable by @​vchulski in #​4476
• Enables continuous language ID for Azure STT by @​MSameerAbbas in #​4479
• fix: OpenAI realtime division by zero by @​darshankparmar in #​4490
• update avatar example and openai readmes by @​tinalenguyen in #​4495
• type cleanup, include all plugins into type checker by @​davidzhao in #​4491
• agents.md and claude.md by @​davidzhao in #​4493
• update examples to use LK Inference by @​davidzhao in #​4494
• #​4500 Fix for Race condition in _send_kill_signal: ValueError: process object is closed after SIGUSR1 by @​rafallezanko in #​4501
• support diarization for soniox stt by @​longcw in #​4510
• fix 11labs tts hang after update_options in tool call by @​longcw in #​4499
• AGT-2316: refine timestamps in spans and recording alignment by @​chenghao-mou in #​4131

New Contributors

• @​wasaybaig201 made their first contribution in #​4365
• @​ChristianBernhard made their first contribution in #​4358
• @​gau-nernst made their first contribution in #​4391
• @​rusg77 made their first contribution in #​4419
• @​lalq made their first contribution in #​4405
• @​guiruggiero made their first contribution in #​4468
• @​vchulski made their first contribution in #​4476
• @​MSameerAbbas made their first contribution in #​4479
• @​rafallezanko made their first contribution in #​4501

Full Changelog: https://github.com/livekit/agents/compare/livekit-agents@1.3.10...livekit-agents@1.3.11

v1.3.10

Compare Source

What's Changed

• fix(google): improve handling of empty responses by @​davidzhao in #​4330
• Add support for audio frame processor by @​lukasIO in #​4145
• Update doc for min_endpointing_delay by @​MonkeyLeeT in #​4327
• force interruption when closing the session by @​longcw in #​4346
• add ProviderTool & support built-in tools for xai & gemini realtime by @​theomonnom in #​4344
• fix dynamic tool updates in llm_node by @​davidzhao in #​4355
• Proper support for V1 models for Google STT by @​chenghao-mou in #​4338
• Add Grok example by @​ShayneP in #​4363
• allow aws realtime to accept str tool results by @​tinalenguyen in #​4364
• (gemini realtime) check for vertexai for api version by @​tinalenguyen in #​4366
• Enable Soniox STT turn detection & metrics by @​matejmarinko-soniox in #​4332

Provider tools

This release brings the ability to use tools that are specific to model providers with provider tools. You can now mix & match function tools and provider tools in your agent by specifying Agent(tools=[..]).

For those that were using the experimental _gemini_tools parameter with Google LLMs, that experimental parameter has been removed in favor of provider tools. See usage example here.

New Contributors

• @​MonkeyLeeT made their first contribution in #​4327

Full Changelog: https://github.com/livekit/agents/compare/livekit-agents@1.3.9...livekit-agents@1.3.10

v1.3.9

Compare Source

What's Changed

• chore(xai): update voice names according to docs by @​davidzhao in #​4295
• Add local dev commands for linking to rtc-sdk by @​lukasIO in #​4258
• Ensure makefile checks for livekit_lib_path by @​lukasIO in #​4298
• feat(gemini3) : Add Gemini 3 support with thinking_level and thought_signature by @​varghesepaul in #​4027
• fix list mutation during iteration by @​theomonnom in #​4304
• add gemini 3 flash model by @​tinalenguyen in #​4301
• Websockets improvement by @​cshape in #​4303
• Allow for Cartesia TTS language auto-detection by @​yuyuma in #​4300
• Add Amazon Nova 2.0 Sonic Support with Text Input and Enhanced Features by @​kachenjr in #​4176
• fix dynamic FieldInfo for pydantic 2.12 by @​longcw in #​4290
• re-export TurnDetection for xAI by @​tinalenguyen in #​4306
• fix commit_user_turn when last_final_transcript_time is None by @​longcw in #​4308
• feat(soniox): add language_hints_strict option for STT by @​cateet in #​4281
• feat(google-tts): add prompt to normal synthesize for Gemini TTS by @​NXV5111 in #​4208
• Adding extra content to OpenAI LLM. Improving function call grouping. by @​russellmartin-livekit in #​4170
• feat(gemini3) use low latency thinking_level by default for gemini 3 models by @​pushkar-nurix in #​4311
• fix handoff to Realtime model with existing session context by @​davidzhao in #​4310
• tts metrics update by @​dhruvladia-sarvam in #​4117
• AGT-2302: add aligned_transcript to STT by @​chenghao-mou in #​4155
• Minor readme doc fixes by @​kachenjr in #​4320
• vad enabled by @​dhruvladia-sarvam in #​4321
• handle exceptions in task_results by @​tinalenguyen in #​4323
• add livekit-durable functions by @​theomonnom in #​4272
• fix py3.10-py3.12 &livekit-durable cibw by @​theomonnom in #​4324
• stringify cartesia error to be pickleable by @​tinalenguyen in #​4328
• livekit-agents 1.3.9 by @​theomonnom in #​4329

New Contributors

• @​varghesepaul made their first contribution in #​4027
• @​NXV5111 made their first contribution in #​4208
• @​russellmartin-livekit made their first contribution in #​4170
• @​dhruvladia-sarvam made their first contribution in #​4117

Full Changelog: https://github.com/livekit/agents/compare/livekit-agents@1.3.8...livekit-agents@1.3.9

v1.3.8

Compare Source

What's Changed

• add init for xai by @​tinalenguyen in #​4286
• fix(xai): list openai as a dependency, fix exports by @​davidzhao in #​4287
• fix(xai): a few more exports by @​davidzhao in #​4288
• Update default model by @​gyang-xai in #​4289
• chore(xai): list supported voices by @​davidzhao in #​4292

New Contributors

• @​gyang-xai made their first contribution in #​4289

Full Changelog: https://github.com/livekit/agents/compare/livekit-agents@1.3.7...livekit-agents@1.3.8

v1.3.7

Compare Source

What's Changed

• fix OTEL types by @​theomonnom in #​4164
• feat(background-audio): add several builtin audio clips by @​rektdeckard in #​4165
• fix gemini function tool parameter enum typing by @​tinalenguyen in #​4166
• use inference gateway in the readme by @​theomonnom in #​3665
• update warm transfer readme and extra instructions by @​longcw in #​4168
• terminate on JobRequest.reject by @​theomonnom in #​4172
• add terminate argument to JobRequest.reject by @​theomonnom in #​4173
• update documentation link for LiveAvatar by @​tinalenguyen in #​4178
• fix logging style format is not respected by @​longcw in #​4169
• fix _on_reject when no answer by @​longcw in #​4180
• expose elevenlabs TTS error message by @​longcw in #​4182
• fix(aws): Handle nested schema in Nova Sonic tool parameter extraction by @​somoore in #​4177
• Restore otel chat message by @​chenghao-mou in #​4118
• fix record.exc_info is not pickable when using LogQueueHandler by @​longcw in #​4185
• Feat/mistralai models update by @​fabitokki in #​4156
• feat(rime): expand update_options to accept all TTS parameters by @​gokuljs in #​4095
• Fallback API for Inference by @​adrian-cowham in #​4099
• Add LiveAvatar Stop Session API Call + README Fix by @​tinalenguyen in #​4195
• feat(google): add streaming support for Gemini TTS models by @​plumber0 in #​4189
• fix watchfiles prevent agent prcoess exit on sigterm by @​longcw in #​4194
• fix race condition when stop background audio play handle by @​longcw in #​4197
• Inference: Rename fallback model name param by @​adrian-cowham in #​4202
• fix inworld punctuation handling by @​cshape in #​4215
• ensure playback_segments_count is consistent in the audio output chain by @​longcw in #​4211
• clear _q_updated right after await to avoid race conditions by @​longcw in #​4209
• fix blocked send task in liveavatar plugin by @​tinalenguyen in #​4214
• feat(warm-transfer): add sip_number parameter for outbound caller ID by @​Hormold in #​4216
• add keep alive task for liveavatar plugin by @​tinalenguyen in #​4231
• turn-detector: remove english model from readme by @​lwestn in #​4233
• feature: GPT-5.2 support by @​pushkar-nurix in #​4235
• disable interruptions for agent greeting by @​hiroshihorie in #​4223
• AGT-2328: negative threshold in silero by @​chenghao-mou in #​4228
• fix: image token usage not being tracked for OpenAI realtime models by @​GigaDroid in #​4238
• check for type key in _ensure_strict_json_schema by @​tinalenguyen in #​4236
• fix(openai): migrate realtime STT to GA API by @​Hormold in #​4232
• fix(google): handle content blocking and generation failures by @​davidzhao in #​4249
• feat(google): update default realtime model to gemini-2.5 12-2025 by [@​davidzhao](https://redirect.github.com/david

---

Configuration

📅 Schedule: Branch creation - At any time (no schedule defined), Automerge - At any time (no schedule defined).

🚦 Automerge: Enabled.

♻ Rebasing: Whenever PR is behind base branch, or you tick the rebase/retry checkbox.

🔕 Ignore: Close this PR and you won't be reminded about these updates again.

---

• If you want to rebase/retry this PR, check this box

---

This PR was generated by Mend Renovate. View the repository job log.