# Fluxions API

> Hosted speech APIs behind one gateway at https://api.fluxions.ai: akro-v1 for transcription (speech-to-text, speaker diarization, non-speech events) and VUI for expressive low-latency text-to-speech over HTTP or WebSocket. Every endpoint takes an API key in the Authorization header.

## API documentation

- [Overview](https://fluxions.ai/docs/01-overview.md): Base URL, API key authentication, and the health check — start here
- [Transcription](https://fluxions.ai/docs/02-transcription.md): akro-v1 speech-to-text with speaker diarization and non-speech events: submit audio, poll for the result
- [Speech](https://fluxions.ai/docs/03-speech.md): VUI text-to-speech over HTTP or a warm WebSocket, with voice cloning and non-verbal cues
- [History](https://fluxions.ai/docs/04-history.md): Read-only history of your transcriptions, renders and voice conversations
- [Errors](https://fluxions.ai/docs/05-errors.md): Every status code we return, the exact error payload, and what to do about it

## Optional

- [Every API page in one file](https://fluxions.ai/docs-combined-api-docs.md): the whole reference, concatenated
- [Pricing](https://fluxions.ai/pricing): per-hour transcription and per-character speech rates