Hushscript turns recorded audio into speaker-labeled transcripts. Everything
the app does with a transcript is also open to scripts and AI agents through
two interfaces that share one account, one minute balance and one bearer
token: the REST API under https://api.hushscript.com/v1/ and the
MCP server.
What the API can do
- Transcribe recordings. Upload prepared audio or import it from a link, then poll the job until the transcript is ready. Video goes in as extracted audio, not as a video file.
- Read and edit transcripts. List, read, export, tag, translate, restore and delete transcripts.
- Insights. Read, generate and delete Insights for a transcript.
- Recognition context. Manage saved dictionaries and prompt presets.
- Account and billing. Read the balance, usage and holds; save a card and buy minute packs when the token and the account allow it.
- Workspaces. Read members, matters and the audit log, and manage member rights and legal holds in a Hushscript for Teams workspace.
- Agent accounts. An agent can open and pay for its own account without a human account holder, then rotate its own credentials.
Each call is limited by the scopes on its token. The full scope table and the endpoint list are on the API page.
Start here
- Developer API: quickstart, authentication, scopes, endpoints, limits and errors.
- SDKs and CLI: official TypeScript, Python, Go, and PHP libraries and a command line tool that transcribe a file in one call.
- MCP server: connect an AI agent to a Hushscript account as tools instead of HTTP calls.
- Automated usage: the end-to-end walkthrough for agents that run their own account.
- OpenAPI document: the
machine-readable reference for every
/v1/operation. - llms.txt: a plain-text summary of the API for language models.