Speechmatics

API

Speech-to-text and text-to-speech API in 55+ languages, cloud or on-prem

Price: Free credits, then paid
Access: API key

About

Speechmatics transcribes recorded files and live audio in 55+ languages and also offers text to speech. New accounts get $100 in credit, and batch transcription starts at $0.129 per hour. On-prem and on-device deployment is for Enterprise plans only.

What you can do with it

Transcribe recorded audio files with speaker diarization
Stream live audio over WebSocket for real-time transcripts
Add speech recognition and speech output to a voice agent

Get started

Create an API key in the Speechmatics portal ($100 free credit)
Pick a regional endpoint such as eu1, us1 or au1

Example

curl -X POST "https://eu1.asr.api.speechmatics.com/v2/jobs/" \
  -H "Authorization: Bearer $SPEECHMATICS_API_KEY" \
  -F data_file=@example.wav \
  -F config='{"type": "transcription", "transcription_config": {"model": "enhanced", "language": "en"}}'
# When the job is done: GET /v2/jobs/<job_id>/transcript?format=txt

Details

Hosting: Hosted service, Self-hosted
Available in: Worldwide
Official SDKs: Python, JavaScript/TypeScript
MCP server: None