wx44wx/pipecat
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity
7,583 Commits 276 Branches 111 Tags
9bb712a47b4ece26dde95fb2b41e8e866c4ed1df
Commit Graph

4386 Commits

Author SHA1 Message Date
Sergio Sillero
8ae834366b Merge branch 'pipecat-ai:main' into feature/genesys_serializer 2026-01-25 21:04:27 +01:00
Mark Backman
a4acc12f91 Align Speechmatics STT TTFB metrics with STT classes 2026-01-24 18:26:34 -05:00
Mark Backman
e93112e76e Simplify STT finalize handling 2026-01-24 15:28:27 -05:00
Mark Backman
d2ac9006a2 Update env var to PIPECAT_SMART_TURN_LOG_DATA 2026-01-24 12:50:42 -05:00
Mark Backman
bcb019e8ab Add TTFB metrics for STT services (#3495) 2026-01-23 18:47:34 -05:00
kompfner
4ea546785f Merge pull request #3406 from omChauhanDev/fix/openrouter-gemini-messages 
fix(openrouter): handle multiple system messages for Gemini models
 2026-01-23 14:53:59 -05:00
filipi87
7921bce4af Refactoring AudioBufferProcessor to fix audio track synchronization. 2026-01-23 16:15:48 -03:00
Luke Payyapilli
cadced3f79 feat: handle server_content.interrupted for faster barge-in response 2026-01-23 10:41:04 -05:00
Akhil
3b3c7aa8cc LLMAssistantAggregator: preserve non-ASCII characters in JSON output 
Add ensure_ascii=False to json.dumps() calls for tool call arguments
and function call results to prevent unnecessary unicode escaping.
 2026-01-22 15:37:44 -06:00
Cale Shapera
6b5bcae86f change default Inworld TTS model to inworld-tts-1.5-max (#3531) 2026-01-22 14:21:15 -05:00
Mark Backman
3d545b718d Merge pull request #3344 from omChauhanDev/fix/stt-dynamic-language-update 
fix: treat language as first-class STT setting
 2026-01-22 09:21:56 -05:00
marcus-daily
76b774072c Formatting fixes 2026-01-22 14:17:59 +00:00
marcus-daily
b6341ffaa5 Save Smart Turn input data if SMART_TURN_LOG_DATA is set 2026-01-22 14:17:59 +00:00
Mark Backman
29fae67c9e Merge pull request #3523 from omChauhanDev/add-location-support-google-tts 
feat(google): add location parameter to TTS services
 2026-01-22 09:12:16 -05:00
Aleix Conchillo Flaqué
de73e28563 Merge pull request #3510 from omChauhanDev/feat/add-reached-filter-methods 
feat(task): add additive filter methods for frame monitoring
 2026-01-21 21:05:33 -08:00
Om Chauhan
7bd32e2fe5 feat(google): add location parameter to TTS services 2026-01-22 09:49:19 +05:30
James Hush
8f05d95f50 feat: add video_out_codec parameter for DailyTransport (#3520) 
* feat: add video_out_codec parameter for DailyTransport

Add video_out_codec parameter to TransportParams allowing configuration
of the preferred video codec (VP8, H264, H265) for video output.

When set, this passes the preferredCodec option to Daily's
VideoPublishingSettings during the join operation.

* chore: move video_out_codec parameter to changelog folder (#3522)

* Initial plan

* Move video_out_codec parameter to changelog/3520.added.md

Co-authored-by: jamsea <614910+jamsea@users.noreply.github.com>

* Revert all CHANGELOG.md changes, keep only changelog/3520.added.md

Co-authored-by: jamsea <614910+jamsea@users.noreply.github.com>

---------

Co-authored-by: copilot-swe-agent[bot] <198982749+Copilot@users.noreply.github.com>
Co-authored-by: jamsea <614910+jamsea@users.noreply.github.com>

---------

Co-authored-by: Copilot <198982749+Copilot@users.noreply.github.com>
Co-authored-by: jamsea <614910+jamsea@users.noreply.github.com>
 2026-01-22 11:31:07 +08:00
Om Chauhan
87c12f3098 changed frame filter storage type from tuples to sets 2026-01-22 08:43:46 +05:30
Om Chauhan
6e44a2ab49 feat(task): add additive filter methods for frame monitoring 2026-01-22 08:43:46 +05:30
Aleix Conchillo Flaqué
0ee11ad333 tests: disable RTVI in tests by default 2026-01-21 18:14:17 -08:00
Aleix Conchillo Flaqué
124a3c35af RTVIObserver: don't handle some frames direction 2026-01-21 18:14:17 -08:00
Aleix Conchillo Flaqué
e85a00cc0e PipelineTask: automatically add RTVI processor and RTVI observer 
If `enable_rtvi` is enabled (enabled by default) and RTVI processor will be
added automatically to the pipeline. Also, and RTVI observer will be
registered.
 2026-01-21 18:14:17 -08:00
Aleix Conchillo Flaqué
cc61cdbba3 RTVIProcessor: add create_rtvi_observer() 2026-01-21 18:14:17 -08:00
Aleix Conchillo Flaqué
62f4708d43 transports: broadcast InputTransportMessageFrame frames 2026-01-21 18:14:17 -08:00
Aleix Conchillo Flaqué
ba0ddb1832 FrameProcessor: copy kwargs when broadcasting frame 2026-01-21 18:14:17 -08:00
Aleix Conchillo Flaqué
eacd2a4b71 FrameProcessor: add broadcast_frame_instance() 2026-01-21 18:14:17 -08:00
okue
4a724379fc refactor(user_mute): remove unnecessary _bot_speaking assignment in _handle_bot_stopped_speaking 
The _bot_speaking flag does not need to be set in this method,
so the redundant assignment has been removed.
 2026-01-21 23:59:15 +09:00
Aleix Conchillo Flaqué
7e0ca113af CambTTSService: initialize client during StartFrame 2026-01-20 19:07:12 -08:00
Aleix Conchillo Flaqué
a787fd9cd8 NVIDIATTSService: process incoming audio frame right away 
Process audio as soon as we receive it from the generator. Previously, we were
reading from the generator and adding elements into a queue until there was no
more data, then we would process the queue.
 2026-01-20 15:41:05 -08:00
Aleix Conchillo Flaqué
14495c425a NVIDIASTTService: no need for additional queue and task 2026-01-20 13:50:17 -08:00
Aleix Conchillo Flaqué
bd45ce2b4e Merge pull request #3499 from lukepayyapilli/fix/livekit-video-queue-memory-leak 
fix(livekit): prevent memory leak when video_in_enabled is False
 2026-01-20 13:21:21 -08:00
Aleix Conchillo Flaqué
a266644b06 Merge pull request #3494 from omChauhanDev/fix/uninterruptible-frame-handling 
fix: preserve UninterruptibleFrames in __reset_process_queue
 2026-01-20 13:19:40 -08:00
Mark Backman
03faadd7f9 Merge pull request #3508 from pipecat-ai/ss/log-daily-ids 
Log Daily participant and meeting session IDs upon successful join in…
 2026-01-20 15:43:48 -05:00
Sunah Suh
fa6f924b31 Log Daily participant and meeting session IDs upon successful join in Daily Transport 2026-01-20 11:31:17 -06:00
Aleix Conchillo Flaqué
655006aff5 NvidiaSegmentedSTTService: simplify exception handling 2026-01-20 08:58:14 -08:00
Aleix Conchillo Flaqué
671dc8cd9b NvidiaSTTService: initialize client on StartFrame 
Initialize client on StartFrame so errrors are reported within the pipeline.
 2026-01-20 08:58:14 -08:00
Aleix Conchillo Flaqué
9a718ded1e NvidiaTTSService: initialize client on StartFrame 
Initialize client on StartFrame so errrors are reported within the pipeline.
 2026-01-20 08:58:14 -08:00
Aleix Conchillo Flaqué
024809b39a Merge pull request #3503 from pipecat-ai/aleix/ai-service-start-end-cancel 
AIService: handle StartFrame/EndFrame/CancelFrame exceptions
 2026-01-20 08:56:39 -08:00
Aleix Conchillo Flaqué
6cf0d53d00 AIService: handle StartFrame/EndFrame/CancelFrame exceptions 
If AIService subclasses implement start()/stop()/cancel() and exception are not
handled, execution will not continue and therefore the originator frames will
not be pushed. This would cause the pipeline to not be started (i.e. StartFrame
would not be pushed downstream) or stopped properly.
 2026-01-20 08:54:22 -08:00
kompfner
778dacc9a8 Merge pull request #3486 from pipecat-ai/pk/fix-nova-sonic-reset-conversation 
Fix `AWSNovaSonicLLMService.reset_conversation()`
 2026-01-20 10:07:38 -05:00
Paul Kompfner
06b3ecd2d6 In AWS Nova Sonic service, send the "interactive" user message (which triggers the bot response) only after sending the audio input start event, per the AWS team's recommendation 2026-01-20 09:56:25 -05:00
ssillerom
fa5da3b0be change comments 2026-01-19 20:49:23 +01:00
ssillerom
7e82a0cf49 feature: Genesys AudioHook WebSocket protocol serializer for Pipecat 2026-01-19 20:45:22 +01:00
Mark Backman
cddd6d5b0a Merge pull request #3492 from pipecat-ai/mb/remove-unused-imports 
Remove unused imports
 2026-01-19 14:07:16 -05:00
Mark Backman
11cf891ac8 Manual updates for unused imports 2026-01-19 14:03:22 -05:00
Luke Payyapilli
dfc1f09b77 fix(livekit): prevent memory leak when video_in_enabled is False 2026-01-19 11:00:23 -05:00
Mark Backman
14bd3b1b32 Set Azure TTS default prosody rate to None 2026-01-19 09:19:57 -05:00
Mark Backman
f733e77496 AzureTTS: work around word ordering issue at 8khz sample rate 2026-01-19 09:13:41 -05:00
Om Chauhan
4a9eb82f92 fix: preserve UninterruptibleFrames in __reset_process_queue 2026-01-18 20:39:13 +05:30
Om Chauhan
990d8386e4 fix: make SmallWebRTCConnection pc_id globally unique 2026-01-18 19:41:51 +05:30