Commit Graph

148 Commits

Author SHA1 Message Date
Paul Kompfner
523e890c8c Fix deprecation messages pointing users to the new import paths for Gemini Live 2025-10-10 10:30:38 -04:00
Aleix Conchillo Flaqué
7cec013666 google: rename google.gemini_live.gemini to google.gemini_live.llm 2025-10-09 22:20:09 -07:00
Aleix Conchillo Flaqué
aeace9b9be google: move gemini_live inside google service 2025-10-09 16:06:42 -07:00
Paul Kompfner
f2d9063984 Renames: remove "multimodal" from Gemini Live types 2025-10-09 10:58:36 -04:00
Paul Kompfner
2699f0c2a6 Fix tool calls when using Gemini Live + Vertex AI 2025-10-08 21:03:07 -04:00
Paul Kompfner
728361a6a7 Add GeminiVertexMultimodalLiveLLMService 2025-10-08 21:03:01 -04:00
Paul Kompfner
cf90071926 Format fix 2025-10-08 17:19:46 -04:00
Paul Kompfner
a666327d70 Implement ending GeminiMultimodalLiveLLMService gracefully (i.e. after the bot is finished) 2025-10-08 17:13:04 -04:00
Paul Kompfner
7da37a0d1f Pull _connection_established_threshold and _max_consecutive_failures into file-level constants 2025-10-08 17:04:05 -04:00
Paul Kompfner
7efb22a323 Add (relatively spartan) reconnection logic to GeminiMultimodalLiveLLMService, leveraging the Gemini Live session resumption mechanism 2025-10-08 16:53:21 -04:00
Paul Kompfner
402e019ae2 Make a bit of code clearer 2025-10-08 16:45:55 -04:00
Paul Kompfner
f09e4e238b Fix some mishandling of enum values 2025-10-08 16:45:55 -04:00
Paul Kompfner
ac1582c906 Let users directly use google-genai types rather than aliased re-exported types 2025-10-08 16:45:55 -04:00
Paul Kompfner
e4b01a5844 Bumping deprecation version of GeminiMultimodalLiveLLMService's base_url arg 2025-10-08 16:45:55 -04:00
Paul Kompfner
d19e6111c3 Bumping deprecation version of GeminiMultimodalLiveLLMService's base_url arg 2025-10-08 16:45:55 -04:00
Paul Kompfner
8a6d504a7e Add enable_affective_dialog and proactivity settings to GeminiMultimodalLiveLLMService 2025-10-08 16:45:55 -04:00
Paul Kompfner
43915937f2 Update how we import and re-export some types in GeminiMultimodalLiveLLMService 2025-10-08 16:45:55 -04:00
Paul Kompfner
48e92a22fe Add thinking settings to GeminiMultimodalLiveLLMService 2025-10-08 16:45:55 -04:00
Paul Kompfner
566af6b0b8 Minor comment improvement 2025-10-08 16:45:55 -04:00
Paul Kompfner
12e7613d5f Deprecate the base_url argument to GeminiMultimodalLiveLLMService.
It expected a WebSocket URL, but we're no longer (directly) using WebSockets to talk to Gemini. Instead of trying to (potentially erroneously) map a given custom WebSocket URL to an `HttpOptions` object (the new preferred way of customizing requests made by the Gemini API client), we're simply deprecating `base_url` and pointing users to the `http_options` argument instead.
2025-10-08 16:45:55 -04:00
Paul Kompfner
04a68f2c57 Fix tracing in GeminiMultimodalLiveLLMService 2025-10-08 16:45:55 -04:00
Paul Kompfner
9b4ca12f49 Revert to only supporting providing a single modality - looks like specifying a list of modalities results in an API error.
Also, fix some missing `await`s in error handling.
2025-10-08 16:45:55 -04:00
Paul Kompfner
453ce715a6 Add some error handling to GeminiMultimodalLiveLLMService 2025-10-08 16:45:55 -04:00
Paul Kompfner
d87b6189ba Update GeminiMultimodalLiveLLMService to use the google-genai library, which is the new recommended approach for interfacing with Gemini Live. 2025-10-08 16:45:55 -04:00
Aleix Conchillo Flaqué
9d9f10ae0e frames: StartInterruptionFrame is deprecated, use InterruptionFrame 2025-09-11 09:01:54 -07:00
Paul Kompfner
246f9f3325 Raise an error when attempting to use Gemini Multimodal Live with universal LLMContext. This is exactly the same error we already have for the other s2s models, AWS Nova Sonic and OpenAI Realtime, it was just missing from this service. 2025-09-04 16:47:08 -04:00
Aleix Conchillo Flaqué
24a628c85e remove watchdog timers and specific asyncio implementations
Watchdog timers have been removed. They were introduced in 0.0.72 to help
diagnose pipeline freezes. Unfortunately, they proved ineffective since they
required developers to use Pipecat-specific queues, iterators, and events to
correctly reset the timer, which limited their usefulness and added friction.
2025-08-21 18:09:56 -07:00
Vanessa Pyne
96b1000e52 Merge pull request #2341 from getchannel/realtime-text
Hotfix: Correct Gemini Live API class to fix 1007 payload error.
2025-08-04 11:03:57 -05:00
Pete
258e83c904 Fix: Correct Gemini Live API text input to prevent 1007 WebSocket errors
- Restore TextInputMessage.realtimeInput structure for correct API format
- Remove invalid turnComplete message from _send_user_text method
- turnComplete is only valid for clientContent, not realtimeInput messages
- realtimeInput text completion is automatically inferred by the API

This fixes WebSocket 1007 errors caused by mixing realtimeInput and
clientContent message types in violation of the Gemini Live API contract.
2025-08-03 10:58:59 -04:00
Vanessa Pyne
af8de227bb Merge pull request #2223 from getchannel/realtime-text
Add text input handling to unify context for realtimeInput stream of GeminiMultimodalLiveService
2025-07-31 08:53:39 -05:00
Pete
2af3b6329d Ruff format debug 2025-07-29 17:48:11 -04:00
Pete
8ca06e5887 Add InputTextRawFrame class for handling raw text input in frames
- Introduced `InputTextRawFrame` to represent raw text input from users or programs.
- Updated `GeminiMultimodalLiveLLMService` to process `InputTextRawFrame` and send user text via the Gemini Live API's realtime input stream.
- Enhanced `_send_user_text` method documentation for clarity on its functionality and usage.
2025-07-29 17:43:14 -04:00
Mark Backman
300f19ad23 Port to the websockets asyncio implementation, support for websockets 13 and 14 2025-07-23 11:54:25 -04:00
Vanessa Pyne
8467d87cfc small main-merge fixes - gemini.py 2025-07-21 09:52:32 -05:00
Pete
b54d1fb7fd Resolve merge conflict and remove duplicate File API initialization
- Remove duplicate file_api initialization lines
- Keep grounding metadata tracking functionality
- Maintain clean code structure
2025-07-20 19:15:40 -04:00
Pete
ec361df0d1 Fix final ruff linting issues
- Remove duplicate import in __init__.py
- Clean up extra blank lines in gemini.py
- Remove extra blank line in _create_single_response method
2025-07-20 18:58:54 -04:00
Pete
b1a5cddde4 Refactor whitespace and formatting in multiple files
- Clean up unnecessary whitespace in `gemini.py`, `events.py`, and `file_api.py`
- Ensure consistent formatting in `26g-gemini-multimodal-live-groundingMetadata.py`
- Improve readability by aligning code and removing trailing spaces
2025-07-20 18:40:12 -04:00
Pete
8ba340a8a5 remove debug logging 2025-07-20 18:21:42 -04:00
Pete
8f74b97591 Refactor _send_user_text method in Gemini multimodal service to streamline event creation for turn completion 2025-07-20 18:08:45 -04:00
Pete
1d69cd1a5e Remove debug logging from _send_user_text method in Gemini multimodal service 2025-07-20 18:04:57 -04:00
Pete
bd7a0f27cc Add text input handling to Gemini multimodal service
- Updated `RealtimeInput` to include an optional `text` parameter.
- Introduced `TextInputMessage` class for encapsulating text input data.
- Implemented `_send_user_text` method to send text input to the Gemini Live API.
- Enhanced message processing to support text input alongside media chunks.
2025-07-20 17:39:31 -04:00
Filipi Fuchter
b573f9dab2 Removing duplicated code inside Gemini. 2025-07-04 10:57:53 -03:00
Pete
7ed4fe50d4 Update gemini.py
-FunctionCallFromLLM
-Delete duplicate Gemini imports
2025-07-03 19:39:44 -04:00
Pete
6f66ec1727 Update gemini.py
tab indentation fix
2025-07-03 18:55:21 -04:00
Pete
c7e758fc36 Merge branch 'main' into groundingMetadata 2025-07-03 18:47:47 -04:00
Pete
4951c97eab Clean up verbose logging in grounding metadata implementation
- Remove debug logging from grounding metadata event handlers
- Simplify logging in _process_grounding_metadata method
- Clean up example file logging for better readability
- Remove verbose event parsing comments

Based on suggestions from draft PR #2121
2025-07-03 17:49:27 -04:00
Vanessa Pyne
d80e228c6f Merge branch 'main' into mcp-http-gemini-support 2025-07-02 11:47:18 -05:00
Mark Backman
f590a476e7 Gemini Live fixes, plus additional docstrings 2025-07-02 08:27:23 -07:00
Yousif Astarabadi
4bcc536fd2 added arg description in docstring for gemini live init 2025-07-02 00:03:27 -07:00
Yousif Astarabadi
1ab2ddd317 fix lint error 2025-07-01 23:55:34 -07:00