Clients

TypeScript SDK

npm install @opentranscription/sdk

Python SDK

pip install opentranscription

CLI

npm install -g @opentranscription/cli

Agent Skill

npx skills add opentranscription/skills

Speech-to-Text API Reference

One REST API for every model in the catalog. Upload your audio, create a job, and read the transcript back in the same shape no matter which provider ran it — switching models is a field in the request, not a rewrite of your integration.

Base URL https://opentranscription.io. Authenticate with a scoped API key sent as a bearer token. Catalog and benchmark reads are public and need no key.

Endpoints

Transcriptions

Create and manage transcription jobs.

POST

/api/v1/transcriptions

Create a transcription job

GET

/api/v1/transcriptions

List transcription jobs

GET

/api/v1/transcriptions/{id}

Get a transcription job

PATCH

/api/v1/transcriptions/{id}

Rename a transcription job

DELETE

/api/v1/transcriptions/{id}

Cancel or delete a transcription job

GET

/api/v1/transcriptions/{id}/audio

Download a transcription job's original audio

POST

/api/v1/uploads

Create a signed upload URL for audio file upload

Usage

Organization usage analytics.

GET

/api/v1/usage

Get usage statistics

Catalog

Public model catalog and benchmark data. No authentication required.

GET

/api/v1/models

List transcription models

GET

/api/v1/models/{modelId}

Get a single model

Benchmarks

Leaderboard and per-model benchmark detail.

GET

/api/v1/benchmarks

Get leaderboard

GET

/api/v1/benchmarks/export

Download leaderboard as CSV

GET

/api/v1/benchmarks/{modelId}

Get model benchmark detail

Service level objectivesVersioning & stability policyOpenAPI spec