gemini-audio

Transcription with timestamps and speaker identification is built into this Skill using Gemini's audio API. Upload your audio file (supports multiple formats, up to 9.5 hours), and the Skill generates a transcript with speaker labels and precise timing for each segment.

Yes. This Skill handles podcasts, meetings, interviews, and any audio content up to 9.5 hours per request. It produces transcripts, summaries, topic extraction, and segment analysis—covering both speech and non-speech sound detection.

Gemini audio transcription supports multiple audio formats through both file uploads and inline data submission. The Skill handles diverse input modalities, processing any format Gemini's API accepts within the 9.5-hour limit per request.

Text-to-speech generation is included via Gemini's TTS capabilities. Provide text input, and the Skill synthesizes audio with controllable voice styles, enabling you to convert written content into spoken audio output.

This Skill supports both gemini-2.5-flash and gemini-2.5-pro models. You can select the model based on your token budget and latency requirements for transcription, analysis, and synthesis tasks.

Each audio request is limited to 9.5 hours maximum duration. The Skill manages this constraint automatically, so longer files must be split into multiple requests—a known edge case for extended recordings or archived content.