Medical speech-to-text

Get the medicine right.

Clinical transcription with the top medical-term accuracy on our sealed benchmark and the strongest dosage score of 30 systems.

Start with 25 free hoursGet the open model
Frankfurt processingSelf-serve BAACustomer data is never used for training
Consultation audio
“Start amoxicillin 500 mg and repeat HbA1c at review.”
✓ Drug and dosage preserved
#1 / 29medical-term accuracy
0.00%drug-name error
97.7%dosage F1 · 86/89 events
Plans

Start free. Scale your way.

Run Omi yourself, use the hosted API, or deploy privately.

Open Source
$0Run it yourself

Deploy omi-medical-edge-1 on-device or in your own cloud.

  • Open weights · CC-BY-4.0
  • Mac, CUDA and CPU
  • Commercial use
  • Audio stays with you
Get the open model
Free API
$025 audio-hours every month

Use the complete hosted API without adding a card.

  • Full transcription API
  • Eight languages
  • Medical vocabulary and speakers
  • Self-serve BAA and DPA
Create a free key
Pro API
$0.20Per audio-hour after the free 25

Keep production running with no subscription or minimum.

  • Your 25 free hours stay free
  • Custom patterns
  • Long-audio flagship profile
  • Higher concurrency
Upgrade to Pro
Enterprise
CustomLicense and deployment

Bring the flagship model or supported SDK into your environment.

  • Private cloud deployment
  • Supported on-device SDK
  • Dedicated capacity
  • SLA and integration support
Talk to us
Benchmark results

Every layer of the medical record, measured.

Overall transcription, medical terminology and dosage accuracy are scored independently on the same clinical audio.

Explore the benchmarkCompare providers

Overall word error · lower is better

Azure5.97
Omi5.99
AWS6.12
ElevenLabs6.17

Medical-term error · lower is better

Omi0.94
ElevenLabs0.97
Google1.11
AssemblyAI1.43

Dosage F1 · higher is better

Omi97.7
Deepgram86.8
ElevenLabs85.4
Azure83.3

Start with your own audio.

Use the hosted API or download the open model.

Start freeGet open source