Speech-to-Text Transcription

To transcribe audio into text or subtitles, you can use this Skill to process supported audio files like mp3, wav, or ogg. It uses 0G Compute providers to convert spoken content into plain text, JSON, or SRT subtitle formats.

Supported audio formats for speech-to-text transcription include mp3, wav, ogg, flac, and webm. The Skill processes these file types through 0G Compute providers to generate accurate text outputs.

Yes, you can generate SRT subtitles from podcast or meeting recordings. The Skill produces multiple output formats including SRT subtitles and verbose JSON with timestamps and segments for spoken-audio workflows.

Yes, authenticated provider access is required to convert audio to text. The Skill handles provider metadata lookup, request signing, and response processing using the @0glabs/0g-serving-broker dependency to manage 0G Compute access.

The best way to get searchable text from recorded interviews is using a transcription Skill that outputs structured JSON. This approach applies format-aware transcription to capture spoken content and convert it into easily searchable text.

Audio transcription can fail due to size, format, or balance issues. The Skill includes operational guardrails that perform basic error checking for these constraints, requiring supported formats like mp3 or wav and adequate provider balance.