fix(stt): dispatch live-API-only Gemini models over the Live WebSocket transport (#4006)

Addresses #4006 by letting Gemini STT models that only exist on the Live
API transcribe instead of failing.

transcribeGemini sends audio to :generateContent and that is the only Gemini
path. A model that is realtime-only (exposed by the Live API's
bidiGenerateContent WebSocket) therefore fails outright, even though the account
can transcribe it.

open-sse/handlers/geminiLiveStt.js owns the WebSocket lifecycle: opens
:bidiGenerateContent, sends setup frame, waits for setupComplete, streams
audio as realtimeInput media chunks, and settles on turnComplete.
Dispatch is driven by transport marker 'gemini-live'. Adds custom model transport
persistence and selection on the dashboard.
This commit is contained in:
Nick Nyanjui
2026-09-26 11:48:48 +07:00
committed by decolua
parent 273f0c32cd
commit fe347e4ea5
10 changed files with 945 additions and 22 deletions

View File

@@ -58,6 +58,7 @@ export default {
{ id: "gemini-2.5-flash", name: "Gemini 2.5 Flash", params: ["language","prompt"], kind: "stt" },
{ id: "gemini-2.5-flash-lite", name: "Gemini 2.5 Flash Lite (Cheapest)", params: ["language","prompt"], kind: "stt" },
{ id: "gemini-2.0-flash", name: "Gemini 2.0 Flash", params: ["language","prompt"], kind: "stt" },
{ id: "gemini-2.5-flash-native-audio-preview-09-17", name: "Gemini Live Transcription (Realtime)", params: ["language","prompt","system_instruction","setup_timeout_ms","turn_timeout_ms"], kind: "stt", transport: "gemini-live" },
{ id: "gemini-3.1-flash-tts-preview", name: "Gemini 3.1 Flash TTS", kind: "tts" },
{ id: "gemini-2.5-flash-preview-tts", name: "Gemini 2.5 Flash TTS", kind: "tts" },
{ id: "gemini-2.5-pro-preview-tts", name: "Gemini 2.5 Pro TTS", kind: "tts" },