Commit Graph

1622 Commits

Author SHA1 Message Date
Aleix Conchillo Flaqué
abb20f34ba Update default Anthropic model to claude-sonnet-4-6
Update the default model in AnthropicLLMService and remove the
now-unnecessary explicit model from the function calling example.
2026-02-20 16:17:51 -08:00
Aleix Conchillo Flaqué
af4ef95dc6 Fix missing await on add_audio_frames_message in Google audio examples
The method is async but was being called without await, silently
discarding the coroutine.
2026-02-20 14:24:22 -08:00
Filipi da Silva Fuchter
c9615c8db6 Merge pull request #3779 from pipecat-ai/filipi/filter_observer
Allowing to define the list of frame processors whose frames should be silently ignored by the RTVI observer.
2026-02-20 12:42:02 -05:00
Mark Backman
82ce3ea8de Update 07c example to use DeepgramSageMakerTTSService 2026-02-20 08:10:41 -07:00
Mark Backman
273692421f Add DeepgramSageMakerTTSService for Deepgram TTS on AWS SageMaker
Adds a TTS service that connects to Deepgram models deployed on AWS
SageMaker endpoints via HTTP/2 bidirectional streaming. Supports the
Deepgram TTS protocol (Speak, Flush, Clear, Close) over the BiDi
client, with interruption handling and per-turn TTFB metrics.

Updates the example and env.example with separate STT/TTS endpoint names.
2026-02-20 08:08:00 -07:00
filipi87
18630c9478 Adding changelog entry for RTVI observer ignored_sources feature. 2026-02-19 18:41:05 -03:00
filipi87
3a8d3cc841 Allowing to define the list of frame processors whose frames should be silently ignored by the RTVI observer. 2026-02-19 18:36:12 -03:00
Mark Backman
507765625f Make UserIdleController always-on with dynamic timeout updates
Always create UserIdleController (timeout=0 means disabled), removing
all Optional guards. Add UserIdleTimeoutUpdateFrame to allow changing
the idle timeout at runtime.
2026-02-14 09:54:30 -05:00
Mark Backman
012ef41ff4 Redesign UserIdleController to use BotStoppedSpeakingFrame
Replace the continuous heartbeat-based timer (UserSpeakingFrame/BotSpeakingFrame
+ asyncio.Event loop) with a simple one-shot timer that starts when
BotStoppedSpeakingFrame is received and cancels on UserStartedSpeakingFrame or
BotStartedSpeakingFrame. This eliminates false idle triggers caused by gaps
between the user finishing speaking and the bot starting to speak (LLM/TTS
latency).

Guard the timer start with two conditions to prevent false triggers:
- User turn in progress: during interruptions, BotStoppedSpeaking arrives
  while the user is still speaking mid-turn.
- Function calls in progress: FunctionCallsStarted arrives before
  BotStoppedSpeaking because the bot speaks concurrently with the function
  call starting, so the timer must wait for the result and subsequent bot
  response.
2026-02-14 08:55:56 -05:00
Mark Backman
18afe37bd1 Add changelog entries for PR #3642 2026-02-12 14:09:24 -05:00
Mark Backman
2b9777b812 Update RimeTTSService InputParams for arcana and mistv2 model support
Add model-specific params (arcana: repetition_penalty, temperature, top_p;
mistv2: no_text_normalization, save_oovs, segment) with dynamic query param
building via _build_settings(). Model/voice/param changes now trigger
WebSocket reconnection since all settings are URL query params.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
2026-02-12 14:01:41 -05:00
Mark Backman
972ad93e18 Fix quickstart pcc-deploy.toml 2026-02-10 22:17:09 -05:00
Mark Backman
b063d9d43b Fix quickstart pyproject.toml 2026-02-10 22:06:38 -05:00
Mark Backman
883b24f577 Update quickstart for 0.0.102 2026-02-10 18:14:04 -05:00
filipi87
ba242d4875 Context summarization example with Google 2026-02-10 18:59:03 -03:00
filipi87
5deb80932b Context summarization example with OpenAI 2026-02-10 18:58:55 -03:00
Filipi da Silva Fuchter
9c627e7292 Merge pull request #3653 from pipecat-ai/filipi/heygen_lite
HeyGen improvements.
2026-02-10 12:12:22 -05:00
Filipi da Silva Fuchter
ad179b0852 Merge pull request #3584 from pipecat-ai/filipi/speak_frame
TTS services improvements.
2026-02-10 12:11:47 -05:00
filipi87
87a79df048 Updating the heygen examples to use sandbox by default. 2026-02-10 14:02:20 -03:00
filipi87
ad1bec4583 Updated openai example to use on_tts_request and append_to_text. 2026-02-10 11:28:35 -03:00
Mark Backman
88e981c013 Set vad_force_turn_endpoint to False in SonioxSTTService 2026-02-10 09:16:03 -05:00
Mark Backman
d47d95e1f0 Update SonioxSTTService default model to stt-rt-v4 2026-02-09 23:48:08 -05:00
Aleix Conchillo Flaqué
83a8379401 examples: remove the now default turn analyzer user turn stop strategy 2026-02-09 14:33:38 -08:00
Mark Backman
5a1deb7cb4 Merge pull request #3659 from pipecat-ai/mb/change-vad-defaults
Set VADParams stop_secs to 0.2 by default
2026-02-06 23:51:50 -05:00
Mark Backman
a5fc2b1650 Set VADParams stop_secs to 0.2 by default 2026-02-06 23:49:08 -05:00
Mark Backman
aa5a855eab Merge pull request #3656 from pipecat-ai/mb/openai-realtime-stt
Add OpenAIRealtimeSTTService
2026-02-06 09:15:58 -05:00
Mark Backman
d11e1cd631 Update 13k to use ElevenLabsRealtimeSTTService 2026-02-05 15:48:00 -05:00
Mark Backman
8b9da632d1 Add OpenAIRealtimeSTTService 2026-02-05 15:48:00 -05:00
Mark Backman
56d8ef2bf4 Deprecate UserBotLatencyLogObserver, update 29 example 2026-02-05 14:29:45 -05:00
Derek Haynes
93138466d6 Feat: Add user-bot latency to OTel turn spans
This adds user-to-bot response latency tracking to OpenTelemetry spans:

- Created UserBotLatencyObserver as a reusable component for tracking
user-to-bot response latency
- Records the value as an attribute on turn spans (turn.user_bot_latency_seconds)
- Updated TurnTraceObserver to use UserBotLatencyObserver, following the same pattern as TurnTrackingObserver
- Updated PipelineTask to automatically create and wire UserBotLatencyObserver
when tracing is enabled (same as TurnTrackingObserver)
2026-02-05 14:29:42 -05:00
Mark Backman
86147f15f3 Renumber the Resemble foundational example 2026-02-02 09:07:05 -05:00
Mark Backman
ba2b7c05d6 Add ResembleAITTSService 2026-02-02 08:55:27 -05:00
Aleix Conchillo Flaqué
675c7c43e3 examples: update 07zd to use vad_analyzer in LLMUserAggregator 2026-01-31 15:31:15 -08:00
Mark Backman
a587e1b99a Update quickstart for 0.0.101 2026-01-31 09:52:24 -05:00
Aleix Conchillo Flaqué
604d5d0b14 examples: update 07zi and 07zj to use vad_analyzer form LLMUserAggregator 2026-01-30 16:14:02 -08:00
Mark Backman
c92ec1552e Add 22 foundational to release evals 2026-01-30 15:12:52 -05:00
Mark Backman
63a23246d5 Add UserTurnCompletionLLMServiceMixin (#3518)
* Added UserTurnCompletionLLMServiceMixin class

* Added 22-filter-incomplete-turns.py foundational example

* Removed old 22 natural conversation foundational examples

* Added test_user_turn_completion_mixin.py
2026-01-30 14:57:15 -05:00
Aleix Conchillo Flaqué
307aeaeda0 examples: update with LLMUserAggregatorParams vad_analyzer and VADProcessor 2026-01-30 10:07:34 -08:00
Aleix Conchillo Flaqué
72ab329513 services(tss): add new KokoroTTSService 2026-01-30 09:39:01 -08:00
Mark Backman
7be2b8cc34 Merge pull request #3587 from pipecat-ai/mb/gradium-improvements
GradiumSTTService now flushes pending transcripts on VAD stopped dete…
2026-01-29 18:11:25 -05:00
Aleix Conchillo Flaqué
bbb8b53d03 runner: allow custom CLI arguments 2026-01-29 10:15:53 -08:00
Mark Backman
31c7fbc5ba Add delay_in_frames and language support 2026-01-29 10:59:04 -05:00
Mark Backman
6ab12626d6 GradiumSTTService now flushes pending transcripts on VAD stopped detection 2026-01-29 10:26:17 -05:00
Aleix Conchillo Flaqué
875614ff7a tts: add support for local PiperTTSService 2026-01-29 00:16:39 -08:00
Aleix Conchillo Flaqué
eb1bf1e446 tts: rename PiperTTSService to PiperHttpTTSService 2026-01-28 23:27:32 -08:00
Filipi da Silva Fuchter
27277ed3d9 Merge pull request #3571 from pipecat-ai/filipi/funcion_call_improvements
Function call improvements
2026-01-28 14:03:40 -05:00
filipi87
c8496dfb8e Updated the examples which use UserImageRequestFrame to defer the function call result. 2026-01-28 15:39:21 -03:00
Aleix Conchillo Flaqué
f3ef488925 rename DAILY_SAMPLE_ROOM_URL to DAILY_ROOM_URL 2026-01-28 10:05:27 -08:00
Mark Backman
f7a1c6b719 Merge pull request #3408 from ai-coustics/aic-v2
Add ai-coustics AIC SDK v2 support with model downloading
2026-01-27 10:38:26 -05:00
Mark Backman
9c6b11cecf Update README links to use absolute URLs 2026-01-26 13:03:39 -05:00