AssemblyAI
Accurate, cheap transcription API with speaker labels and summaries.
Free
- Best for
- Developers transcribing many hours
- On your PC
- No, online only
- Works on
- API, Web
- Made by
- AssemblyAI
What’s good
- Very low price for this accuracy
- Fourth on the multilingual Open ASR leaderboard
- Speaker labels, summaries and live streaming built in
Downsides
- Made for developers, not a point-and-click app
- Streaming is billed while the connection is open
At a glance.
- Best for
- Developers transcribing many hours
- Price
- Free
- For developers
- $0.21 per hour (Universal-3.5 Pro); streaming from $0.15 per hour
- Version
- Universal-3.5 Pro
- Works on
- API, Web
- Open source
- No
- Made by
- AssemblyAI
- Website
- assemblyai.com
How it ranks.
- Open ASR Leaderboard multilingual, average WER#4 · 4.482026-11-05
- Open ASR Leaderboard English, average WER#11 · 6.232026-11-05
- Artificial Analysis speech to text, AA-WER %#15 · 32026-11-05
Can I run it myself?
No. It only runs online.
Your prompts and files go to the company’s servers. See the local options in this category for a private alternative.
Want it offline? The best local option for speech to text is Whisper large-v3 turbo.