Free APIVision & SpeechAPI keyOfficial
Deepgram
$200 in non-expiring credits, no card, ~430 hours of Nova speech-to-text before you pay
Deepgram is a speech AI platform offering fast, accurate speech-to-text (Nova models), text-to-speech and a voice agent API. The single /v1/listen endpoint handles both pre-recorded files and remote URLs with features like diarization, punctuation, smart formatting and language detection. The free credit applies across STT, TTS and audio intelligence.
Free tier
$200 free credit, no card, credits do not expire; then pay-as-you-go (Nova-3 from ~$0.0048/min batch)Rate limits
Pay-as-you-go accounts have limited concurrency (fine for dev, a bottleneck at scale); exact limits not publishedAuth
API keyBase URL
https://api.deepgram.comWhat you'd build with it
- Transcribe meeting or podcast audio inside an automation
- Build a real-time voice agent step that streams audio over WebSocket
- Add diarized captions to user-uploaded video
Key endpoints
- POST
/v1/listenTranscribe a remote-URL or uploaded pre-recorded audio file - POST
/v1/speakSynthesize speech (text-to-speech)
Members
Members get the runnable recipe
Sign in free to copy the example request, see a sample response, and read the gotchas.
Tags
speech-to-texttext-to-speechaudiotranscriptionvoice