SpragDocs
Speech to text, text to speech, realtime voice, and voice creation over one OpenAI-compatible HTTP API.
Speech to text
Transcribe recorded audio over REST, or stream a live transcript over a realtime WebSocket.
Get startedText to speech
Synthesize speech from text with preset, designed, or cloned voices.
Get startedSprag serves speech models over an HTTP API that follows the OpenAI shape. To move code that already talks to an OpenAI-compatible endpoint onto Sprag, change the base URL and the API key.
Use speech to text in one of two forms: a REST endpoint for audio that already exists, or a realtime WebSocket for audio that is still arriving. Use text to speech to speak a string you send over REST, or open a realtime session in which the model hears the user and speaks its own reply.
Base URL
https://api.sprag.ai/v1Before you begin
Create a Sprag account and generate a key from API keys in the app. Then read the quickstart for authentication, rate limits, and a request you can paste into a terminal.
Where to go next
Integrations. Run Pipecat agents on Sprag
models with the pipecat-sprag plugin, and LiveKit agents by pointing
LiveKit's OpenAI plugin at the Sprag base URL.
API reference. Look up request and response schemas for the REST endpoints. For the realtime WebSocket, see realtime transcription and realtime speech.