ElevenLabs: Scribe v2 Medical vs Mistral: Voxtral Mini Transcribe
A side-by-side comparison of ElevenLabs: Scribe v2 Medical ($0.01 / minute) and Mistral: Voxtral Mini Transcribe ($0.01 / minute) using current USD prices and capabilities from the api14 catalog. Both run through the same OpenAI-compatible API.
Specifications
| ElevenLabs: Scribe v2 Medical | Mistral: Voxtral Mini Transcribe | |
|---|---|---|
| Provider | ElevenLabs | Mistral |
| Audio | $0.01 / minute | $0.01 / minute |
| Context window | - | 16,384 |
| Max output | - | 13,107 |
| Input | audio | audio |
| Output | transcription | transcription |
| Tool calling | ||
| Structured output | ||
| Reasoning control | ||
| Added | Oct 7, 2026 | May 15, 2026 |
Key differences
- Mistral: Voxtral Mini Transcribe is 18% cheaper on input tokens.
In depth
ElevenLabs: Scribe v2 Medical is a transcription model from ElevenLabs, listed on api14 under the model id elevenlabs/scribe-v2-medical. Mistral: Voxtral Mini Transcribe is a transcription model from Mistral, listed under the model id mistralai/voxtral-mini-transcribe. Both sit in the transcription category, so they solve the same basic problem: you send audio in and get a transcription out. Neither is a general chat or reasoning model, and neither is meant to generate text from a text prompt. The choice between them comes down to fit for your specific audio workload rather than to a difference in what kind of product they are.
On capabilities the two are close. Both accept audio as their only input modality and both return a transcription as their only output modality. Neither exposes any supported parameters, so you cannot tune sampling, temperature or similar knobs through the API for either one. The clearest listed difference is context: Voxtral Mini Transcribe has a context length of 16384 tokens, while Scribe v2 Medical has no context length listed. The qualitative comparison marks biggerContext as null, so there is no stated basis for claiming one handles longer audio than the other. The other qualitative fields, including tools, structured output and reasoning, are also null, meaning neither model is flagged as uniquely offering those. The one directional difference given is price: Voxtral Mini Transcribe is the cheaper input side, while the output side is a tie. So if input cost dominates your bill, Voxtral Mini Transcribe is the cheaper of the two, though the page's live price table is the place to check the actual numbers.
For practical selection, start from the domain. Scribe v2 Medical is positioned by its name and provider around medical transcription, so if your audio is clinical in nature, that is the natural first call to evaluate. Voxtral Mini Transcribe is the more general option and is the cheaper input side, which makes it a reasonable default when you are transcribing large volumes of ordinary audio and want to keep input spend down. Because the listed capability fields do not separate them on tools, structured output or reasoning, you should not pick one over the other expecting a feature the other lacks. Instead, test both on a sample of your own audio and compare the transcripts directly, since that is the difference that will actually matter to your users.
Calling either model through api14 works the same way because the API is OpenAI-compatible. You point your existing client at the api14 base URL, send your audio in the request, and set the model field to the id you want. To switch between them you change only that one string, from elevenlabs/scribe-v2-medical to mistralai/voxtral-mini-transcribe or back. Your request shape, authentication and response handling stay as they are, so you can run both against the same audio and compare results without rewriting integration code.
There is a case where the other model is still the better pick. If your audio is not medical and your priority is the lowest input cost, Voxtral Mini Transcribe is the cheaper input side and the sensible choice even though Scribe v2 Medical exists. Conversely, if your audio is medical, Scribe v2 Medical is the one aligned with that use case, and paying the higher input rate may be justified. Since the output side is a tie and no other capability separates them, that domain fit plus the input price difference is the whole decision.
Switch with one string
Both models use the same endpoint and request shape. Change the model value from "elevenlabs/scribe-v2-medical" to "mistralai/voxtral-mini-transcribe" and prompts, streaming and tools keep working.
import fs from "node:fs";
import OpenAI from "openai";
const client = new OpenAI({
baseURL: "https://api14.com/api/v1",
apiKey: process.env.API14_API_KEY,
});
const result = await client.audio.transcriptions.create({
model: "mistralai/voxtral-mini-transcribe",
file: fs.createReadStream("audio.mp3"),
});
console.log(result.text);
More comparisons
- AssemblyAI: Universal-3.6 Pro vs Mistral: Voxtral Mini Transcribe 3
- AssemblyAI: Universal-3.6 Pro vs Mistral: Voxtral Small Transcribe 3
- AssemblyAI: Universal-3.6 Pro vs Moondream: Parakeet Ultra
- AssemblyAI: Universal-3.6 Pro vs NVIDIA: Nemotron 3.5 Transcribe
- Mistral: Voxtral Mini Transcribe 3 vs Mistral: Voxtral Small Transcribe 3
- Mistral: Voxtral Mini Transcribe 3 vs Moondream: Parakeet Ultra
- Mistral: Voxtral Mini Transcribe 3 vs NVIDIA: Nemotron 3.5 Transcribe
- Mistral: Voxtral Small Transcribe 3 vs Moondream: Parakeet Ultra