Open source

Run medical transcription on your hardware

Omi Med STT v1 is a separate open-source, English medical transcription runtime. Audio stays in your process; no Omi API key or hosted account is required.

Install and transcribe

# Base CLI
pip install -U omi-med-stt

# Apple Silicon / MLX
pip install -U "omi-med-stt[mlx]"
omi-med-stt consultation.m4a --json

# Verify the local runtime and model artifact
omi-med-stt check

The package reads common audio formats, downmixes to mono, and resamples to 16 kHz. Model weights download on first use from the published Hugging Face repositories.

Choose a runtime

HardwareRuntimeReleased artifact
Apple SiliconmlxMLX q8
NVIDIA CUDAGPU runtimeCanonical published weights
Linux / Windows CPUcppGGUF q8_0

Runtime exports have separate benchmark rows because quantization and conversion can change accuracy. Use the result for the artifact you actually deploy, not a different published artifact's number. See GitHub for the current platform-specific GPU installation command.

Cloud and local are different products

CapabilityHosted APIOpen-source runtime
ProcessingManaged in the EUInside your process
LanguagesEight supported families; English medically evaluatedEnglish
Long jobs and webhooksYesYour application owns the workflow
Speaker labels and word timingAvailable by request fieldsNot part of the released CLI contract
Encounter vocabulary and PatternsAvailable by request fields and planNot part of the released CLI contract
LicenseService termsMIT runtime code; CC-BY-4.0 weights

Review transcripts before clinical use

The local runtime is speech-to-text, not a diagnostic, triage, prescribing, or clinical decision system. It is not clinically validated. Keep human review in the workflow.