← 返回广场
查看 X 原帖
使用 Gemini Interactions API 从 URL 转录音频
作者: Philipp Schmid模型: nanoBanana-Pro发布时间: 2026/2/5 16:49:31
类目
标签
#API Integration#Speech Recognition#Audio Transcription#Timestamp Detection#Speaker Identification
复制提示词
Did you know you can transcribe audio directly from a url using the Gemini Interactions API? Gemini 3 Flash can detect timestamps and separate speakers. Prompt: ``` Listen to this meeting recording. Transcribe the speech and identify the speakers as Speaker A, Speaker B, etc. Output the result in a format: [Time] [Speaker]: [Text] ```