26 Releases published by 2 people

• [email protected]

  published May 15, 2025

• [email protected]

  published May 17, 2025

• [email protected]

  published May 18, 2025

• [email protected]

  published May 18, 2025

• [email protected]

  published May 18, 2025

• [email protected]

  published May 18, 2025

• [email protected]

  published May 18, 2025

• [email protected]

  published May 18, 2025

• [email protected]

  published May 18, 2025

• [email protected]

  published May 18, 2025

• [email protected]

  published May 18, 2025

• [email protected]

  published May 18, 2025

• [email protected]

  published May 18, 2025

• [email protected]

  published May 18, 2025

• [email protected]

  published May 18, 2025

• [email protected]

  published May 18, 2025

• [email protected]

  published May 18, 2025

• [email protected]

  published May 18, 2025

• [email protected]

  published May 18, 2025

• [email protected]

  published May 18, 2025

• [email protected]

  published May 18, 2025

• [email protected]

  published May 18, 2025

• [email protected]

  published May 18, 2025

• [email protected]

  published May 18, 2025

• [email protected]

  published May 18, 2025

• [email protected]

  published May 18, 2025

32 Pull requests merged by 12 people

• Version Packages

  #2323 merged May 18, 2025

• update livekit-protocol and livekit-api dependencies to match 1.0

  #2315 merged May 18, 2025

• fix readme example links

  #2319 merged May 17, 2025

• v1.0.22

  #2320 merged May 17, 2025

• updates to readme

  #2318 merged May 17, 2025

• update text-only example to show generate_reply usage

  #2317 merged May 17, 2025

• openai: pass tool choice to session initialization

  #2316 merged May 17, 2025

• added deepwiki badge for weekly repo refresh

  #2296 merged May 16, 2025

• Add support for decoding preconnect buffer

  #2311 merged May 16, 2025

• push a silent audio frame when input stream closed

  #2309 merged May 16, 2025

• avatar runner publish audio/video tracks until first frame pushed

  #2308 merged May 16, 2025

• fix AudioByteStream for multi channels

  #2307 merged May 16, 2025

• aws: fixed STT error at end of session

  #2297 merged May 16, 2025

• openai: switch to mp3 encoding by default

  #2306 merged May 16, 2025

• streaming for google tts

  #2143 merged May 16, 2025

• Update README.md

  #2299 merged May 16, 2025

• fix user transcription forwarding order

  #2279 merged May 15, 2025

• add pre-connected audio buffer

  #2171 merged May 15, 2025

• v1.0.21

  #2294 merged May 15, 2025

• feat: allow to switch between Gemini and Vertex AI using env vars

  #2292 merged May 14, 2025

• Fix Neuphonic CI Tests.

  #2287 merged May 14, 2025

• Partial fix for Neuphonic bug detailed in issue #2281.

  #2282 merged May 14, 2025

• correctly pass timeout as connect timeout instead of total

  #2285 merged May 14, 2025

• fix gather not retrieved error in stt

  #2276 merged May 14, 2025

• feat: Add max_completion_tokens parameter to OpenAI LLM

  #2258 merged May 14, 2025

• revert buffer changes to AudioStreamDecoder

  #2286 merged May 14, 2025

• add realtime model metrics

  #2275 merged May 14, 2025

• deepgram: default to aura-2 for TTS

  #2272 merged May 13, 2025

• remove leading whitespace from gemini user transcripts

  #2254 merged May 13, 2025

• support simple word and sentence split for character-based language

  #2263 merged May 12, 2025

• fix raw function tool for google gemini

  #2270 merged May 12, 2025

• fix google-cloud-texttospeech and anthropic min versions

  #2268 merged May 12, 2025

7 Pull requests opened by 5 people

• feat(plugin-meta): add meta plugin for agents

  #2283 opened May 13, 2025

• add call_later_if_silent to wait for a silence period

  #2289 opened May 14, 2025

• Expose google live api input config

  #2291 opened May 14, 2025

• add to_provider_format for ChatContext

  #2295 opened May 15, 2025

• [feature] Add support for input_audio_noise_reduction in RealtimeModel (for OpenAI Realtime API)

  #2313 opened May 16, 2025

• fix chat context item order

  #2321 opened May 17, 2025

• Expose all realtime model parameters

  #2324 opened May 19, 2025

29 Issues closed by 12 people

• RoomComposite Egress only records the first room, subsequent rooms are not recorded

  #2298 closed May 18, 2025

• Feature Request: Add Dynamic Control for Chat Input and Interruption Settings During Session

  #2168 closed May 16, 2025

• Google TTS has streaming disabled

  #2284 closed May 16, 2025

• AWS STT (Transcribe) Timeout

  #2099 closed May 16, 2025

• text chat input is not working

  #2303 closed May 15, 2025

• weather_agent.py sample fails with aws plugins

  #2203 closed May 15, 2025

• Support strict openai function calls

  #1312 closed May 15, 2025

• Neuphonic and Cartesia websockets raise an error after long periods of silence

  #2281 closed May 15, 2025

• False Interim Transcriptions Segments Bug Causing Broken Chat Items

  #2278 closed May 15, 2025

• Using livekit.agents.voice_assistant and it shows up as modulenotfound

  #2290 closed May 14, 2025

• lk.transcription_final attribute is never "true" in the TTS transcription

  #2288 closed May 14, 2025

• GatheringFuture CancelledError in Deepgram STT when using Push-to-Talk

  #2271 closed May 14, 2025

• WebSocket closes with code 1011 when using google.beta.realtime.RealtimeModel

  #2274 closed May 14, 2025

• How to demo in Streamlit?

  #2239 closed May 14, 2025

• TTS voice cracking with background music after v1.0.18 update

  #2277 closed May 14, 2025

• The agent saying words like: "tools_output" during function-call - gemini real time API

  #2174 closed May 14, 2025

• 'error': 401 - Access denied

  #2188 closed May 14, 2025

• Realtime Multimodal metrics and transcription callbacks not firing in `[email protected]`

  #2222 closed May 14, 2025

• Missing Instruction option in OpenAI Realtime

  #2273 closed May 13, 2025

• LLMMetrics Not Emitted When Using LLMFallbackAdapter in LiveKit Agent

  #2269 closed May 13, 2025

• Stub file not found for "livekit.plugins"PylancereportMissingTypeStubs

  #531 closed May 13, 2025

• Memory and Connection Issue

  #2228 closed May 13, 2025

• Gemini sometimes responds to its own question

  #1684 closed May 13, 2025

• "ImportError: 'livekit.agents.pipeline' module not found in voice assistant project"

  #2264 closed May 13, 2025

• Invalid Response Status from Deepgram TTS API

  #2224 closed May 13, 2025

• Chinese Response Streaming Issue in Structured Output Example

  #1985 closed May 12, 2025

• Error when using Gemini model with newly added native MCP support

  #2266 closed May 12, 2025

• Facing issue on adding anthropic and google plugins.Iam using version 1.0.20 for both.

  #2267 closed May 12, 2025

• Audio/Text forwarding cancellation handling

  #2265 closed May 12, 2025

10 Issues opened by 10 people

• Google TTS always fails for first call

  #2322 opened May 18, 2025

• Agent 1.0 voice agent does not work properly with some of ARM64 CPUs

  #2314 opened May 16, 2025

• Add support for `input_audio_noise_reduction` ( `near_field` / `far_field` ) in `RealtimeModel`

  #2312 opened May 16, 2025

• Feature Request: Recording Audio Inside The Agent

  #2310 opened May 16, 2025

• Azure STT fails with error: SpeechRecognition session stopped

  #2305 opened May 16, 2025

• Tool Schema Specified with `@function_tool(raw_schema={<SCHEMA>})` Broken

  #2304 opened May 16, 2025

• Tool failed to initialise with complex nullable params

  #2302 opened May 15, 2025

• Deepgram STT latency measurement

  #2301 opened May 15, 2025

• Ground and elastic search for Gemini LLM and Gemini live

  #2300 opened May 15, 2025

• Allow configuration of `ensure_ascii` in JSON logging for non-English characters

  #2280 opened May 13, 2025

13 Unresolved conversations

Sometimes conversations happen on old items that aren’t yet closed. Here is a list of all the Issues and Pull Requests with unresolved conversations.

• feat(plugin-openai): add support for meta llama

  #2175 commented on May 15, 2025 • 4 new comments

• feat: Add `with_letta` OpenAI plugin

  #2182 commented on May 16, 2025 • 2 new comments

• How to set outbound proxy in livekit

  #2252 commented on May 12, 2025 • 0 new comments

• Agent transcription missing in conversation_item_added

  #2216 commented on May 12, 2025 • 0 new comments

• Custom Plugin for STT || Arabic

  #2262 commented on May 12, 2025 • 0 new comments

• feat: add opentelemetry instrumentation for tracing

  #2260 commented on May 13, 2025 • 0 new comments

• Additional data on .say()

  #2212 commented on May 13, 2025 • 0 new comments

• TTSSegmentsForwarder ERROR- agent stops speaking after this for 2-3 seconds. Facing this issue randomly.

  #797 commented on May 15, 2025 • 0 new comments

• Does LiveKit Support OpenTelemetry or Do I Need to Implement My Own Tracing?

  #2232 commented on May 16, 2025 • 0 new comments

• Google gemini-2.5-flash responses can throw several errors that are fatal to the worker

  #2133 commented on May 18, 2025 • 0 new comments

• feat(whisper): add Whisper plugin for LiveKit

  #1495 commented on May 16, 2025 • 0 new comments

• examples for 1.0

  #1532 commented on May 16, 2025 • 0 new comments

• [draft] add hedra plugin

  #2163 commented on May 16, 2025 • 0 new comments