# Speechall

Speechall is a unified speech-to-text API. It lets developers and agents transcribe audio with models from multiple providers through one authenticated API and one billing account.

## When to use Speechall

Use Speechall when a task needs a transcript, subtitles (SRT or VTT), speaker diarization, language-aware transcription, or a comparison of available speech-to-text models. It is a good fit for processing local recordings, remotely hosted audio, interviews, meetings, podcasts, support calls, and video captions. Use the CLI for one-off or shell workflows; use the REST API for application integrations.

## Start here

- API documentation: https://speechall.com/docs
- Developer resource index: https://speechall.com/developers
- Canonical OpenAPI specification: https://speechall.com/api/openapi.yaml
- API discovery endpoint: https://speechall.com/api/status
- API health check: https://api.speechall.com/
- API base URL: https://api.speechall.com/v1
- Authentication guide and API keys: https://speechall.com/docs/getting-started
- Official CLI: https://speechall.com/cli
- Sitemap: https://speechall.com/sitemap.xml

## How an agent should call the API

Create an API key in the Speechall console, then send it as `Authorization: Bearer YOUR_API_KEY`. Discover models with `GET /speech-to-text-models`. Submit audio to `POST /transcribe` or use `POST /transcribe-remote` for a public audio URL. Every operation, parameter, response, and error schema is documented in the canonical OpenAPI specification.

## CLI

Install the official CLI with `brew install Speechall/tap/speechall`. Set `SPEECHALL_API_KEY`, then run `speechall recording.wav --output-format json`. Run `speechall models` to retrieve the current machine-readable model catalog.

## Trust and support

Speechall is a product of Actondon UG (haftungsbeschränkt). For support or partnership questions, visit https://speechall.com/contact or email info@actondon.com. Read the privacy policy at https://speechall.com/privacy-policy.