pipecat Releases
156 releases of pipecat-ai/pipecat
- 5.2.0
This release introduces the ability to place headers and footers in fixed positions and adds support for using a template engine.
Nov 11, 2024
- v0.0.48
This release adds input queues for frame processors, audio filters, output transport mixers, and various new services.
Nov 10, 2024
- 5.1.6
This release includes a fix to set read-only status for MVT fields based on variable presence and resolves a CSS error in the air-datepicker component.
Oct 26, 2024
- v0.0.47
This release adds AssemblyAISTTService and Gladia transcription support with API updates.
Oct 22, 2024
- v0.0.46
This release adds audio passthrough for STT services, PlayHT service updates, and renames OpenAILLMServiceRealtimeBeta.
Oct 20, 2024
- v0.0.45
This release moves metrics messages from transport's base output to RTVI.
Oct 16, 2024
- v0.0.44
This release adds OpenAI Realtime API support, RTVI bot transcription processor, AudioBufferProcessor, and CanonicalMetricsService.
Oct 16, 2024
- v0.0.43
Added a Markdown text filter for TTS services and an RTVI user LLM text processor. Refactored transport message urgency handling and improved TTS language conversion.
Oct 10, 2024
- v0.0.42
Major update introducing new TTS services (Google, AWS Polly, Lmnt), LivekitTransport, and pipeline clocks. Implemented frame processor tasks and AsyncGeneratorProcessor.
Oct 2, 2024
- v0.0.41
Added LivekitFrameSerializer and fixed bugs in FastAPI and Anthropic services. Corrected errors in the studypal example.
Sep 10, 2024
- v0.0.40
Added RTVI 0.1 protocol support, FatalErrorFrame, and GStreamer pipeline support. Enabled usage metrics and added control frames for bot speaking states.
Aug 20, 2024
- v0.0.39
Fixed a regression introduced in version 0.0.38 that was causing the Pipeline to stop when using Daily transcription.
Jul 23, 2024
- v0.0.38
Added validation and caching options for Silero VAD, and fixed initial metrics formatting. Improved STT time formats.
Jul 23, 2024
- v0.0.37
Implemented RTVIProcessor and added control frames for interrupting the bot and updating context/voice. Removed legacy LLM response frames.
Jul 23, 2024
- v0.0.36
Added Gladia and Coqui X (XTTS) STT services. Introduced idle processors and AsyncFrameProcessor for handling asynchronous services like Deepgram.
Jul 2, 2024
- v0.0.35
Updated API requirements for FastAPI and Twilio transports to enforce serialization parameters.
Jun 28, 2024
- v0.0.34
Fixed issues with asynchronous STT services (Deepgram and Azure) causing transcriptions to be ignored, and corrected an LLM output length regression.
Jun 26, 2024
- v0.0.33
Upgraded to Cartesia's new Python library 1.0.0, fixed asynchronous STT issues with interruptions, and resolved a blocking situation in the Daily transport.
Jun 25, 2024
- v0.0.32
Added new transports for FastAPI and Twilio, support for Azure Speech-To-Text, and improved performance by converting transports to fully use asyncio.
Jun 22, 2024
- v0.0.31
Optimized performance by adjusting the audio frame chunk size to improve processing efficiency.
Jun 13, 2024