Commit Graph

  • 6ebfea4746 update numba version pin to >= Matt 2026-02-27 12:44:31 -05:00
  • f74af9b9c7 Always apply a timeout to summarization LLM calls Mark Backman 2026-02-27 11:57:22 -05:00
  • 82c249608f Move dedicated LLM summarization into LLMContextSummarizer Mark Backman 2026-02-27 11:47:55 -05:00
  • 98e737b4e9 Add tests for context summarization improvements Mark Backman 2026-02-26 22:56:10 -05:00
  • ec9ddb3199 Add changelog entries for context summarization improvements (#3855) Mark Backman 2026-02-26 22:09:26 -05:00
  • 712305c5b1 Add example 54c showing custom context summarization Mark Backman 2026-02-26 21:45:22 -05:00
  • be8ea818c8 Add on_summary_applied event for observability Mark Backman 2026-02-26 21:34:01 -05:00
  • 50710e9c3f Add summarization timeout to prevent hung LLM calls Mark Backman 2026-02-26 21:31:04 -05:00
  • a489bfaf00 Add optional dedicated LLM for context summarization Mark Backman 2026-02-26 21:22:52 -05:00
  • 945a523eed Add configurable summary_message_template to LLMContextSummarizationConfig Mark Backman 2026-02-26 21:03:05 -05:00
  • 790c434a08 Update summary message role: use user instead of assistant Mark Backman 2026-02-26 20:58:06 -05:00
  • db40a354be Merge pull request #3794 from omChauhanDev/fix/context-summarization-llm-specific-message Filipi da Silva Fuchter 2026-02-27 10:57:34 -05:00
  • aa6d3b38b3 Add explanatory comments for LLMSpecificMessage guards in context summarization, amd fixed the missing guard in LLMContextSummarizer._apply_summary when searching for the first system message. filipi87 2026-02-27 12:53:25 -03:00
  • 41d6470e4a Fix docs workflow: add auto-docs label, remove version info Mark Backman 2026-02-27 10:39:37 -05:00
  • 601822e3e5 Add changelog for PR #3865 Mark Backman 2026-02-27 10:25:48 -05:00
  • 3a32d91c66 Set finalized flag on ElevenLabs Realtime STT transcriptions for manual commit strategy Mark Backman 2026-02-27 10:21:10 -05:00
  • 35b3803ebc Merge pull request #3845 from pipecat-ai/filipi/fix_tts_speak_frame Filipi da Silva Fuchter 2026-02-27 09:59:33 -05:00
  • 3b427a47b6 Fixing Piper test. filipi87 2026-02-27 11:57:11 -03:00
  • d701c3427c Changelog entry for the TTSSpeakFrame fix. filipi87 2026-02-27 11:57:03 -03:00
  • 1f45e80f9d Updated the 52-live-translation.py example to demonstrate the fix filipi87 2026-02-27 11:56:52 -03:00
  • bc6f8e51de Fixed TTSSpeakFrame not automatically committing spoken text to the conversation context when used outside of an LLM response (e.g., for bot greeting messages or injected speech) filipi87 2026-02-27 11:56:44 -03:00
  • deba2515f9 Added a new LLMAssistantPushAggregationFrame control frame that signals LLMAssistantAggregator to immediately flush its text buffer to the conversation context filipi87 2026-02-27 11:56:36 -03:00
  • 127b52bad5 Merge pull request #3852 from pipecat-ai/mb/deprecate-processing-metrics Mark Backman 2026-02-27 09:50:29 -05:00
  • 0697f72dae Merge pull request #3864 from pipecat-ai/mb/auto-docs-update Mark Backman 2026-02-27 09:36:27 -05:00
  • c259a6a73b Deprecate processing metrics (ProcessingMetricsData) Mark Backman 2026-02-26 19:52:34 -05:00
  • 3e04f5d05f Add GitHub Actions workflow to auto-update docs on PR merge Mark Backman 2026-02-27 09:18:15 -05:00
  • cd07937c5d Fix missing imports: Add UserStartedSpeakingFrame and UserStoppedSpeakingFrame zack 2026-02-26 22:18:02 -05:00
  • 72934bd8ae Add u3-rt-pro support and improvements to AssemblyAI STT service zack 2026-02-26 21:35:47 -05:00
  • 2a6a993869 Merge pull request #3850 from rupesh-svg/fix/genesys-remove-audio-chunk-logging Mark Backman 2026-02-26 21:52:54 -05:00
  • 4b4e8b839c Add changelog for PR #3851 mb/remove-processing-metrics Mark Backman 2026-02-26 18:27:50 -05:00
  • 86c2dd5cfc Remove processing metrics (ProcessingMetricsData) Mark Backman 2026-02-26 18:20:49 -05:00
  • bbaa79fef0 Add changelog for PR #3850 Rupesh 2026-02-26 14:00:34 -08:00
  • fff9db0d8f Remove verbose audio chunk logging from GenesysAudioHookSerializer Rupesh 2026-02-26 13:51:05 -08:00
  • 7fe458fe59 Merge pull request #3817 from pipecat-ai/pk/service-settings-fix-back-compat-for-nested-external-sdk-types kompfner 2026-02-26 11:08:27 -05:00
  • faed775d90 Extract _DeepgramSTTSettingsBase with shared _merge_live_options_delta to deduplicate LiveOptions merge logic between __init__ and apply_update, and between the Deepgram STT and SageMaker variants; make top-level model/language take precedence over conflicting live_options values in updates; remove unnecessary Language enum-to-string conversion (Language is a StrEnum) Paul Kompfner 2026-02-26 11:02:44 -05:00
  • b63ca524f5 Merge pull request #3806 from pipecat-ai/mb/ultravox-updates Mark Backman 2026-02-26 10:49:21 -05:00
  • 907ff58d41 Align Ultravox Realtime service with OpenAI/Gemini patterns Mark Backman 2026-02-23 17:11:08 -05:00
  • 97b93ebe57 Merge pull request #3696 from pipecat-ai/mb/streaming-tts-input Mark Backman 2026-02-26 10:26:53 -05:00
  • 3ae173520e Code review feedback Mark Backman 2026-02-26 08:59:38 -05:00
  • c184ac09b8 Inline _build_live_options into _connect in DeepgramSTTService and DeepgramSageMakerSTTService since it's trivial and only called from one place Paul Kompfner 2026-02-26 09:42:15 -05:00
  • 3c20eda8bf Keep model/language in LiveOptions at construction time so apply_update's bidirectional sync is sufficient; simplify _build_live_options to only add sample_rate Paul Kompfner 2026-02-26 09:32:52 -05:00
  • d69a337def Add text_aggregation_mode parameter to TTSService Mark Backman 2026-02-10 13:55:19 -05:00
  • f7434cdde1 Add text aggregation time metric for TTS sentence aggregation Mark Backman 2026-02-10 13:54:57 -05:00
  • e21e8585f0 Add deepgram and sagemaker extras to CI test dependencies so Deepgram and Deepgram Sagemaker settings tests can run Paul Kompfner 2026-02-25 18:40:44 -05:00
  • 8b6aa4b912 Unflatten LiveOptions back into a single live_options field on DeepgramSTTSettings and DeepgramSageMakerSTTSettings; add apply_update override with delta-merge semantics and from_mapping override for backward-compatible dict-style updates Paul Kompfner 2026-02-25 18:18:47 -05:00
  • a4b6db6fb4 Flatten LiveOptions into individual fields on DeepgramSTTSettings and DeepgramSageMakerSTTSettings for backward-compatible dict-style updates via STTUpdateSettingsFrame; during the big service settings refactor, we accidentally got rid of the ability to update individual LiveOptions fields with a sparse update Paul Kompfner 2026-02-24 15:06:06 -05:00
  • edc79d374a Merge pull request #3836 from pipecat-ai/mb/small-webrtc-prebuilt-2.3.0 Mark Backman 2026-02-25 17:18:32 -05:00
  • e521aef5df Merge pull request #3842 from pipecat-ai/mb/claude-plugin-docs Mark Backman 2026-02-25 16:38:16 -05:00
  • 3cfff51205 Merge pull request #3827 from pipecat-ai/pk/gemini-tts-service-remove-model-ivar kompfner 2026-02-25 16:14:38 -05:00
  • 3d8e3a4043 Remove unnecessary _model ivar from ElevenLabs STT services, using _settings.model instead. Paul Kompfner 2026-02-24 21:49:54 -05:00
  • 7ee0400c4c Remove unnecessary _model ivar from Hathora TTS and STT services, using _settings.model instead. Paul Kompfner 2026-02-24 21:22:44 -05:00
  • 781d191509 Remove unnecessary _model ivar from GeminiTTSService, using _settings.model instead Paul Kompfner 2026-02-24 21:19:39 -05:00
  • a8cb2a26d1 Merge pull request #3841 from pipecat-ai/pk/groq-tweaks kompfner 2026-02-25 15:54:33 -05:00
  • b1df1ba5d4 Merge pull request #3834 from pipecat-ai/pk/make-ai-service-exclusive-syncer-of-model-name-to-metrics kompfner 2026-02-25 15:53:59 -05:00
  • eee2ef7e85 Add /update-docs skill to claude-plugin Mark Backman 2026-02-25 15:45:16 -05:00
  • ff0f3dce32 A few Groq-related tweaks: - Wire up passing speed setting to Groq, even though only a value of 1.0 is supported today - Update the 55y example to switch voices instead of changing speed - Add a 55zn example to exercise runtime updates of Groq STT Paul Kompfner 2026-02-25 15:10:11 -05:00
  • bca42f7d68 Fix Hathora 55 series examples, and fix Hathora missing settings field warning Paul Kompfner 2026-02-25 14:03:53 -05:00
  • 27940d83a2 Make it so that AIService is the exclusive "syncer" of model name to metrics. Paul Kompfner 2026-02-25 09:41:23 -05:00
  • 937c691f2a Merge pull request #3838 from pipecat-ai/mb/remove-playht Mark Backman 2026-02-25 14:34:15 -05:00
  • 6803d38d3f Merge pull request #3833 from pipecat-ai/mb/add-performance-changelog-fragment Mark Backman 2026-02-25 14:33:52 -05:00
  • 44993fe9e3 Remove PlayHT TTS services Mark Backman 2026-02-25 14:09:17 -05:00
  • 0fe4c732b7 Merge pull request #3837 from alts/alts/append-trailing-space Mark Backman 2026-02-25 13:35:07 -05:00
  • ceead60ef2 Add append_trailing_space to all Rime websocket services Stephen Altamirano 2026-02-25 09:43:57 -08:00
  • e028194dbe Update the pipecat-ai-small-webrtc-prebuilt to 2.3.0 Mark Backman 2026-02-25 12:20:21 -05:00
  • 81f4672535 Add Performance as a changelog fragment option Mark Backman 2026-02-25 09:47:42 -05:00
  • 9273b158ea Merge pull request #3825 from pipecat-ai/mb/llm-user-aggregator-interim-transcription Mark Backman 2026-02-25 09:06:34 -05:00
  • 353a28842c Merge pull request #3807 from pipecat-ai/mb/update-openai-realtime-1.5 Mark Backman 2026-02-25 09:06:19 -05:00
  • 3e6c59c736 Merge pull request #3809 from pipecat-ai/mb/krisp-viva-result Mark Backman 2026-02-25 09:05:12 -05:00
  • 0ca8c850fb Add TurnMetricsData and e2e processing time for KrispVivaTurn Mark Backman 2026-02-24 17:42:18 -05:00
  • 73ee4da7d4 Add Krisp API key support for new SDK licensing requirement Mark Backman 2026-02-24 10:16:35 -05:00
  • 2f60074da3 Merge pull request #3814 from pipecat-ai/filipi/fix_close_context Filipi da Silva Fuchter 2026-02-25 08:21:04 -05:00
  • 751b1b8100 Adding the changelog entries for the tts fixes. filipi87 2026-02-25 10:18:25 -03:00
  • d899f0af11 Refactored all AudioContextTTSService based providers to override the new callbacks instead of _handle_interruption(), making provider-specific cleanup cleaner and more explicit filipi87 2026-02-25 10:18:16 -03:00
  • c09ae6ba6d Added two new lifecycle callbacks to AudioContextTTSService: on_audio_context_interrupted() and on_audio_context_completed() filipi87 2026-02-25 10:17:54 -03:00
  • a187a4b3b2 Merge pull request #3830 from pipecat-ai/aleix/restore-dev-skills Mark Backman 2026-02-25 06:33:16 -05:00
  • 68e19a730b Restore dev skills and add marketplace for maintainer workflows Aleix Conchillo Flaqué 2026-02-24 23:47:06 -08:00
  • 67cb7d575f Merge pull request #3828 from pipecat-ai/mb/skip-empty-audio-filter-frames Mark Backman 2026-02-24 23:27:22 -05:00
  • a84930dc3e Skip empty audio frames after filter buffering Mark Backman 2026-02-24 23:21:52 -05:00
  • 54fd73c460 Merge pull request #3821 from pipecat-ai/pk/fix-missing-field-warning-rime-tts kompfner 2026-02-24 21:58:17 -05:00
  • c954e1c898 Merge pull request #3820 from pipecat-ai/pk/fix-breakage-when-sending-generic-settings-update kompfner 2026-02-24 21:58:05 -05:00
  • db76cd052a Merge pull request #3819 from pipecat-ai/pk/make-update-settings-frames-uninterruptible kompfner 2026-02-24 21:57:41 -05:00
  • 167e68672b Add changelog for InterimTranscriptionFrame/TranslationFrame fix Mark Backman 2026-02-24 20:52:16 -05:00
  • 69d916ca51 Consume InterimTranscriptionFrame and TranslationFrame in LLMUserAggregator Mark Backman 2026-02-24 20:51:41 -05:00
  • 9890b93d08 Merge pull request #3822 from pipecat-ai/fix/stt-ttfb-timeout-timestamp Mark Backman 2026-02-24 19:45:15 -05:00
  • f928206b3a Add changelog for STT TTFB timeout fix Mark Backman 2026-02-24 19:02:40 -05:00
  • f421ad9cf6 Fix STT TTFB timeout measuring to timeout expiry instead of transcript time Mark Backman 2026-02-24 18:57:38 -05:00
  • d918a20b75 Fix missing field warning in RimeTTSService Paul Kompfner 2026-02-24 18:14:16 -05:00
  • d91c230b85 Fix breakage when using a generic settings update (e.g. a TTSSettings) instead of a more specific one (e.g. a RimeTTSSettings). Both should work, assuming you're only changing fields present in the generic settings. Paul Kompfner 2026-02-24 18:05:27 -05:00
  • b6f21ab15d Make ServiceUpdateSettingsFrame uninterruptible—settings updates are generally independent of specific utterances. Paul Kompfner 2026-02-24 17:07:20 -05:00
  • 6f0061ab96 Merge pull request #3812 from pipecat-ai/pk/service-settings-storage-v-delta-mode kompfner 2026-02-24 15:37:49 -05:00
  • 761397d1f9 Merge pull request #3816 from pipecat-ai/aleix/use-skills-repo Aleix Conchillo Flaqué 2026-02-24 12:11:20 -08:00
  • d9bb4d07c6 Merge pull request #3815 from pipecat-ai/fix/sentry-metrics-signatures Aleix Conchillo Flaqué 2026-02-24 12:11:07 -08:00
  • ee46cbce4c Move skills to pipecat-ai/skills repo, add README instructions Aleix Conchillo Flaqué 2026-02-24 11:38:04 -08:00
  • b4b9976b9c Fix SentryMetrics method signatures to match base class Aleix Conchillo Flaqué 2026-02-24 11:26:34 -08:00
  • b78a293ffb Flatten input_params into individual fields on SonioxSTTSettings and GladiaSTTSettings Paul Kompfner 2026-02-24 12:20:14 -05:00
  • 0a89d24f70 Update some more services to ensure that there are no un-initialized fields in self._settings Paul Kompfner 2026-02-24 12:00:23 -05:00
  • 8c9ccf8f82 Bump various deprecation messages from mentioning version 0.0.103 to 0.0.104 Paul Kompfner 2026-02-24 11:47:29 -05:00
  • bcc2b4def4 Make clearer the distinction between "storage-mode" and "delta-mode" usage of *Settings objects - Storage mode: for use in self._settings. All fields should be specified, i.e. should not be NOT_GIVEN. - Delta mode: for use in *UpdateSettingsFrame. Paul Kompfner 2026-02-24 10:52:05 -05:00
  • 57d25c564c Merge pull request #3786 from pipecat-ai/filipi/refactor_word_tts_service Filipi da Silva Fuchter 2026-02-24 13:53:58 -05:00
  • 6cda2ff941 Changelog entry for word timestamp refactor and deprecation notes. filipi87 2026-02-24 15:49:02 -03:00