Speech 🔈

End-to-End blazingly fast streamable Speech-to-Text and Speech Translation with Speaker Diarization and OpenAI compatible.


Able to transcribe Standard Malay, local Malay, Standard English, Manglish, Mandarin, Indonesian and Tamil as audio input.

USA map

Better Accuracy

🔈 Base achieved 100% on prepared Hallucination-Free Test, mesolitica/speech-test-set/hallucination-free, higher is better.

We benchmarked on Malaysian Speech Translation and Speech-to-Text test set, mesolitica/malaysian-stt-leaderboard, higher is better.

Developer Playground

You can play around with Speech 🔈 at Dashboard

  • 160 TPS 🔈 Base, 280 TPS 🔈 Small
  • Real-time SRT format
  • Record Mode and Upload Mode
  • File Upload, accept MP3, WAV or FLAC, max up to 30MB
  • Real-time Speaker Diarization
  • Generate code for OpenAI Python, OpenAI NodeJS, Python AIOHTTP Streaming and CuRL

Prepaid pricing

Best either for solo or a team.

  • Speech Translate to Malay or English
  • Max 30 MB upload
  • Real-time Speaker Diarization
  • Streaming capability

🔈 Base

per 1 Hour

🔈 Small

per 1 Hour

Frequently asked questions

How does Speech API charged if I have a long audio?

We charged every 30 successful processed seconds.

What is the rate limit?

Currently we hard limit 2 hours per Minute.

How to topup?

Just go to billing page and topup! Minimum RM3 and Maximum RM99.

Interested for Enterprise solution?

If you are interested to self-host in your virtual private network either on-premise or private cloud with custom solution, email us at khalil@mesolitica.com or husein@mesolitica.com