--- title: Medical Speech-to-Text API Now Available for Telnyx Voice AI canonical: https://telnyx.com/release-notes/omi-health-medical-stt-voice-ai --- # Medical Speech-to-Text API Now Available for Telnyx Voice AI [Release notes](https://telnyx.com/release-notes) # Medical Speech-to-Text Now Available for Telnyx Voice AI ### 22, Sep 2026 Omi Health's medical speech-to-text model, `omi-health/omi-med-stt-v1`, is now available as a Telnyx STT engine, bringing self-hosted, medical-tuned transcription to voice AI workflows. Built as a fine-tune of `nvidia/parakeet-tdt-0.6b-v2`, the model targets clinical dialogue, medication names, and dosages, and runs on the same self-hosted Parakeet stack as `nvidia/parakeet-v3`. ### What's new * **Medical-tuned STT model:** `omi-health/omi-med-stt-v1` is trained on clinical dialogue and transcribes medication names, dosages, and medical terminology more accurately than general-purpose models. * **Self-hosted on Telnyx infrastructure:** Runs on the Parakeet self-hosted stack, with no third-party API hop for transcription traffic. * **Available on WebSocket, Voice API, TeXML, AI Assistants, and the REST API:** Select the model on Call Control `transcription_start`, TeXML `` or ``, WebSocket streaming, AI Assistants transcription settings, or the Speech-to-Text REST API for file transcription. * **English only:** No language parameter needed since the model is pinned to English. On streaming surfaces it returns final transcripts only, no interim results, and ignores endpointing, and it accepts `linear16` or `linear32` audio at 16 kHz. ### Why it matters * Clinical intake, medication review, and post-visit documentation workflows need accurate transcription of drug names and medical terminology that general-purpose STT models often get wrong. * Self-hosted on Telnyx infrastructure, so audio never leaves the Telnyx network to reach a third-party STT vendor, relevant for healthcare voice AI handling patient conversations. * Adds a medical-domain option to the STT lineup and reuses the same Call Control, WebSocket, TeXML, AI Assistants, and REST config every other engine already uses, so there's no new integration to build. * On-network processing keeps audio and inference on the same private backbone, removing a vendor hop and the latency it adds. ### Example use cases * Telehealth intake lines transcribing patient-reported symptoms, medications, and dosages for clinical documentation. * Pharmacy or provider callback lines that need accurate drug-name transcription without routing patient audio to a third-party vendor. * Clinical dictation and post-visit summary pipelines that need a medical-tuned transcript before it reaches an EHR workflow. * Voice AI assistants fielding medication-review or symptom-triage calls where terminology accuracy affects the response. ### Getting started 1. In Mission Control, go to **AI > Assistants** and open the assistant you want to configure. 2. Select the **Voice** tab and scroll to **Transcription**. 3. Set the transcription engine to **Parakeet** and the model to `omi-health/omi-med-stt-v1`. 4. Save the assistant and test with clinical audio before routing production traffic. For Call Control, POST to `transcription_start`: { "transcription_engine": "Parakeet", "transcription_engine_config": { "transcription_engine": "Parakeet", "transcription_model": "omi-health/omi-med-stt-v1" } } For TeXML ``: Please describe your symptoms after the tone. For WebSocket streaming: wss://api.telnyx.com/v2/speech-to-text/transcription?transcription_engine=Parakeet&model=omi-health/omi-med-stt-v1&input_format=linear16&sample_rate=16000 On streaming surfaces, `omi-health/omi-med-stt-v1` accepts `linear16` or `linear32` audio at 16 kHz and returns final transcripts only. No interim or partial results. Endpointing is ignored. For file transcription, POST to the Speech-to-Text REST API. It's OpenAI SDK compatible, so swapping `base_url` and `api_key` is enough for existing Whisper-style code: from openai import OpenAI client = OpenAI( api_key="YOUR_TELNYX_API_KEY", base_url="https://api.telnyx.com/v2", ) result = client.audio.transcriptions.create( model="omi-health/omi-med-stt-v1", file=open("visit-recording.mp3", "rb"), ) The REST endpoint transcribes synchronously, forces mono, and caps uploads at 100 MB. **Learn more** in the [Speech-to-Text models docs](https://developers.telnyx.com/docs/voice/stt/models), the [Speech-to-Text REST API overview](https://developers.telnyx.com/docs/voice/stt/rest-api), or the [AI Assistants transcription settings](https://developers.telnyx.com/docs/inference/ai-assistants/transcription-settings). [Start building](https://telnyx.com/sign-up) ### Ask AI * [![GPT](https://d2efw2akbx8gaf.cloudfront.net/https://dotcom-payload-assets-prod.s3.us-east-1.amazonaws.com/cms-media/chatgpt_icon-ea55e09d.jpeg?width=92&format=webp)](https://chat.openai.com/?q=What makes Telnyx's global network different, breakdown call quality for normal phone calls and voice AI calls.) * [![Claude](https://d2efw2akbx8gaf.cloudfront.net/https://dotcom-payload-assets-prod.s3.us-east-1.amazonaws.com/cms-media/claude_icon-c7c8eef3.webp?width=92&format=webp)](https://claude.ai/new?q=What makes Telnyx's global network different, breakdown call quality for normal phone calls and voice AI calls.) * [![Perplexity](https://d2efw2akbx8gaf.cloudfront.net/https://dotcom-payload-assets-prod.s3.us-east-1.amazonaws.com/cms-media/perplexity_icon-cd5b74f1.webp?width=92&format=webp)](https://www.perplexity.ai/search?q=What makes Telnyx's global network different, breakdown call quality for normal phone calls and voice AI calls.) * [![Gemini](https://d2efw2akbx8gaf.cloudfront.net/https://dotcom-payload-assets-prod.s3.us-east-1.amazonaws.com/cms-media/gemini_icon-4ff1cbf4.webp?width=92&format=webp)](https://www.google.com/search?udm=50&aep=11&q=What makes Telnyx's global network different, breakdown call quality for normal phone calls and voice AI calls.) * [![Grok](https://d2efw2akbx8gaf.cloudfront.net/https://dotcom-payload-assets-prod.s3.us-east-1.amazonaws.com/cms-media/grok_icon-e588e5bd.webp?width=92&format=webp)](https://x.com/i/grok?text=What makes Telnyx's global network different, breakdown call quality for normal phone calls and voice AI calls.)