谷歌发布 Gemini 3.5 语音转录 API
推荐指数 71.0 NO. 013 · 2026.08.28
发布2026/08/27Score149Comments34
为什么值得看
Gemini 3.5 Transcribe 是谷歌新一代语音转录模型,直接处理原始音频输出格式化文本,已面向开发者开放 API。对需要高精度语音交互的 AI 应用(如会议记录、客服、实时字幕)是重要基础设施升级。
编辑判断
语音转录赛道长期被 Whisper 统治,但开发者的真实痛点不是准确率数字,而是口语中的嗯啊停顿、多人重叠对话、专业术语识别这些工程细节。谷歌这次把 Transcribe 直接塞进 Gemini 3.5 的 API 体系,意味着语音能力可以和 Gemini 的多模态推理、工具调用打通,这是 Whisper 做不到的架构优势。
已经在用 Whisper API 的团队建议做一轮对比测试,特别是长音频和嘈杂场景。另外注意定价策略——谷歌通常会以低于 OpenAI 30-50% 的价格切入企业市场,成本敏感项目可以观望两周后的社区 benchmark。
社区反馈
意见分歧 30 条评论
核心争论:Gemini 3.5 转录精度领先但存在语义简化问题,需权衡便利性与准确性
相关内容
谷歌發布Gemini 3.5 Transcribe語音轉文字模型 Gemini 3.5 Transcribe能將音訊轉換為格式化文字,串流模式詞錯誤率4.0%,非串流2.6%,較前代Chirp 3縮短70%轉錄時間,提供Live API與Interactions API兩種接口。 谷歌推出迄今「最精準」語音轉文本模型,從「逐字記錄」邁向「理解式轉寫」 Gemini 3.5 Transcribe已通過Gemini API進入預覽階段,支援實時語音流與錄製音頻,最長支援1小時文件轉錄,已落地Gboard Rambler及Mac版Gemini應用,計劃引入Chrome瀏覽器。 Gemini 3.5 Audio (Live Translate) - Model Card Gemini 3.5 Transcribe Live通過Gemini API、Gemini App、Google AI Studio、Google Cloud/Vertex AI等渠道分發,涵蓋開發者與企業級應用場景。
I'd love to know how this handles proper subtitle formatting. I'm in the process of learning many languages, and being able to cross check my own understanding with film and video would be fantastic.
That's really impressive! It says it's available in GBoard on Android but I don't see it yet. I assume like most Google features it will be a rolling release over the course of several months, can't wait to try it some day.
It is confusing. As far as I can tell "rambler for Android" is a named subfeature of gboard and only on Pixel 10+ and newer Samsung phones. From a user standpoint it "just" seems like the transcription has gotten better - not an app you use or a feature that you enable.