Unverified50% confidenceFactExact time
Gemini 3.5 Live Translate 同时暴露了输入音频的转录（input transcription）接口
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
8/7/2026
First Seen
Valid until: 11/5/2026
Sources
Gemini 3.5实时字幕实测：电竞解说场景下的语音转录表现
Related Claims
UnverifiedGemini 3.5 Live Translate encapsulates the four steps of speech recognition, semantic understanding, translation, and speech synthesis into a single multimodal large model for joint optimization.79% similarVerifiedGoogle released Gemini 3.5 Live Translate, built specifically for real-time speech-to-speech translation.75% similarUnverifiedGemini's speech generation model is currently based on Gemini 2.5 and supports multilingual and multi-accent output.73% similarVerifiedGemini 3.5 Pro supports five modality inputs: text, images, video, audio, and code70% similarUnverifiedWhen Gemini 3.1 Pro generates videos, the generated dialogue carries a 'Chinese accent.'69% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/705481API
curl https://kongchang.com/api/v1/knowledge/claims/705481MCP
get_claim(id=705481)