Mistral’s Voxtral goes past transcription with summarization, speech-triggered capabilities

Mistral launched an open-sourced voice mannequin at the moment that would rival paid voice AI, corresponding to these from ElevenLabs and Hume AI, which the corporate mentioned bridges the hole between proprietary speech recognition fashions and the extra open, but error-prone variations. 

Voxtral, which Mistral will launch below an Apache 2.0 license, is out there in a 24B parameter model and a 3B variant. The bigger mannequin is meant for functions at scale, whereas the smaller model would work for native and edge use instances. 

“Voice was humanity’s first interface—lengthy earlier than writing or typing, it allow us to share concepts, coordinate work, and construct relationships. As digital techniques grow to be extra succesful, voice is returning as our most pure type of human-computer interplay,” Mistral mentioned in a weblog submit. “But at the moment’s techniques stay restricted—unreliable, proprietary, and too brittle for real-world use. Closing this hole calls for instruments with distinctive transcription, deep understanding, multilingual fluency, and open, versatile deployment.”

Voxtral is out there on Mistral’s API and a transcription-only endpoint on its web site. The fashions are additionally accessible by way of Le Chat, Mistral’s chat platform. 

Mistral mentioned that speech AI “meant selecting between two trade-offs,” stating that some open-source automated speech recognition fashions usually had restricted semantic understanding. Nonetheless, closed fashions with robust language understanding come at a excessive price. 

Bridging the hole

The corporate mentioned Voxtral “provides state-of-the-art accuracy and native semantic understanding within the open, at lower than half the value of comparable APIs.” 

Voxtral, at a 32K token context, can hearken to and transcribe as much as half-hour of audio or 40 minutes of audio understanding. It provides summarization, that means the mannequin can reply questions primarily based on the audio content material and generate summaries with out switching to a separate mode. Customers can set off capabilities and API calls primarily based on spoken directions.

The mannequin relies on Mistral’s Mistral Small 3.1. It helps a number of languages and might mechanically detect languages corresponding to English, Spanish, French, Portuguese, Hindi, German, Italian, and Dutch. 

Mistral added enterprise options to Voxtral, together with non-public deployment, in order that organizations can combine the mannequin into their very own ecosystems. These options additionally embrace domain-specific fine-tuning and superior context and precedence entry to engineering assets for patrons who need assistance integrating Voxtral into their workflows. 

Efficiency 

Speech recognition AI is now out there on many platforms at the moment. Customers can communicate to ChatGPT, and the platform will course of spoken directions equally to written prompts. Quick meals chains like White Fortress have deployed SoundHound to their drive-thru companies, and ElevenLabs has steadily been enhancing its multimodal platform. The open-source area additionally provides highly effective choices. Nari Labs, a startup, launched the open-source speech mannequin Dia in April. Nevertheless, a few of these companies may be fairly costly.

Transcription companies like Otter and Learn.ai can now embed themselves into Zoom conferences, recording, summarizing and even alerting customers to actionable gadgets. Many on-line video assembly platforms supply not simply transcription, but in addition speech AI and agentic AI, with Google Conferences offering the choice to take notes for customers utilizing Gemini. As an everyday consumer of voice transcription companies, I can say firsthand that speech recognition AI just isn’t good, however it’s enhancing.

Mistral acknowledged that Voxtral outperformed present voice fashions, together with OpenAI’s Whisper, Gemini 2.5 Flash and Scribe from ElevenLabs. Voxtral offered fewer phrase errors in comparison with Whisper, which is at the moment thought-about the very best automated speech recognition mannequin out there. 

When it comes to audio understanding, Voxtral Small is “aggressive with GPT-4o-mini and Gemini 2.5 Flash throughout all duties, attaining state-of-the-art efficiency in Speech Translation.”

Since asserting Voxtral, social media customers mentioned they’ve been ready for an open-source speech mannequin that may match the efficiency of Whisper. 

See more

Sure! We wanted this. Every week in the past, I used to be lamenting over a closed-source AI universe and cyberpunk dystopian future, however at the moment, with this addition, my outlook is way improved – go open-source. https://t.co/QsKAfTOxou

— David Hendrickson (@TeksEdge) July 15, 2025

Mistral mentioned Voxtral can be out there by way of its API at $0.001 per minute.