Mistral Unveils Groundbreaking Open-Source Audio AI Model Voxtral
As AI systems grow more sophisticated, speech is rapidly emerging as the primary way we interact with machines. French AI startup Mistral has entered the audio arena with its first open model, challenging the dominance of closed corporate systems by offering open-weight alternatives.
On Tuesday, Mistral introduced Voxtral, its inaugural family of audio models designed for business use.
The company positions Voxtral as the first open model capable of delivering "truly usable speech intelligence in production."
This means developers no longer face a choice between an affordable but inaccurate open system that struggles with transcriptions and lacks real comprehension, and a functional but closed system that comes with higher costs and limited deployment control.
For businesses, Voxtral presents a cost-effective alternative that Mistral claims is "less than half the price" of comparable solutions.

Image Credits: Mistral Mistral states that Voxtral can transcribe up to 30 minutes of audio. Thanks to its LLM backbone, Mistral Small 3.1, it comprehends up to 40 minutes, enabling users to ask questions about the audio, generate summaries, or convert voice commands into real-time actions like API calls or function execution. Voxtral is multilingual, capable of transcribing and understanding languages including English, Spanish, French, Portuguese, Hindi, German, Dutch, and Italian.
The company is releasing two variants of its "speech understanding models." The first, Voxtral Small, features 24B parameters for production-scale deployments and competes with ElevenLabs Scribe, GPT-4o-mini, and Gemini 2.5 Flash.
Techcrunch event LIVE NOW! TechCrunch All Stage
Build smarter. Scale faster. Connect deeper. Join leaders from Precursor Ventures, NEA, Index Ventures, Underscore VC, and more for a day filled with strategies, workshops, and valuable networking.
Save $450 on your TechCrunch All Stage pass
Build smarter. Scale faster. Connect deeper. Join leaders from Precursor Ventures, NEA, Index Ventures, Underscore VC, and more for a day filled with strategies, workshops, and valuable networking.
Boston, MA | July 15 REGISTER NOW The second, Voxtral Mini, has 3 billion parameters for local and edge deployments. There's also an ultra-affordable, streamlined, fast API version of the 3B model named Voxtral Mini Transcribe, optimized solely for transcription tasks and designed to outperform OpenAI Whisper at less than half the cost.
Users can test Voxtral for free by downloading the API from Hugging Face or trying the models in Mistral's chatbot Le Chat. API integration for applications starts at $0.001 per minute, according to the company.
This release follows Mistral's announcement of Magistral last month, its first family of reasoning models that solve problems step-by-step for enhanced reliability.
Mistral, a leading AI firm in Europe, is renowned for its advocacy of open-source AI models. Earlier this month, TechCrunch reported that the company is negotiating to raise up to $1 billion in equity from investors such as Abu Dhabi's MGX fund.
Related article
Google Tests Remy AI Agent for Gemini as Focus Shifts to User Control
According to Business Insider, Google is testing Remy, a new AI personal agent for Gemini. This tool aims to execute tasks on behalf of users, streamlining both professional workflows and daily routines.Currently, Remy is undergoing testing in an int
Ollie bets privacy focus to win AI assistant race
To be genuinely helpful, an AI assistant must understand its user deeply. Ollie, a personal assistant designed for daily life, operates on the premise that this doesn’t require surrendering your data or compromising your privacy.While certain enterpr
How AI LIVE: London Will Explore AI & Industrial Automation
The summit will convene C-suite executives from around the globe to address pressing challenges in global industries, ranging from AI-driven disruption to economic volatility.AI LIVE: The London Summit will gather over 2,000 international leaders und
Related Special Topic Recommendations
Comments (1)
0/500
As AI systems grow more sophisticated, speech is rapidly emerging as the primary way we interact with machines. French AI startup Mistral has entered the audio arena with its first open model, challenging the dominance of closed corporate systems by offering open-weight alternatives.
On Tuesday, Mistral introduced Voxtral, its inaugural family of audio models designed for business use.
The company positions Voxtral as the first open model capable of delivering "truly usable speech intelligence in production."
This means developers no longer face a choice between an affordable but inaccurate open system that struggles with transcriptions and lacks real comprehension, and a functional but closed system that comes with higher costs and limited deployment control.
For businesses, Voxtral presents a cost-effective alternative that Mistral claims is "less than half the price" of comparable solutions.

Mistral states that Voxtral can transcribe up to 30 minutes of audio. Thanks to its LLM backbone, Mistral Small 3.1, it comprehends up to 40 minutes, enabling users to ask questions about the audio, generate summaries, or convert voice commands into real-time actions like API calls or function execution. Voxtral is multilingual, capable of transcribing and understanding languages including English, Spanish, French, Portuguese, Hindi, German, Dutch, and Italian.
The company is releasing two variants of its "speech understanding models." The first, Voxtral Small, features 24B parameters for production-scale deployments and competes with ElevenLabs Scribe, GPT-4o-mini, and Gemini 2.5 Flash.
Techcrunch eventLIVE NOW! TechCrunch All Stage
Build smarter. Scale faster. Connect deeper. Join leaders from Precursor Ventures, NEA, Index Ventures, Underscore VC, and more for a day filled with strategies, workshops, and valuable networking.
Save $450 on your TechCrunch All Stage pass
Build smarter. Scale faster. Connect deeper. Join leaders from Precursor Ventures, NEA, Index Ventures, Underscore VC, and more for a day filled with strategies, workshops, and valuable networking.
Boston, MA | July 15 REGISTER NOWThe second, Voxtral Mini, has 3 billion parameters for local and edge deployments. There's also an ultra-affordable, streamlined, fast API version of the 3B model named Voxtral Mini Transcribe, optimized solely for transcription tasks and designed to outperform OpenAI Whisper at less than half the cost.
Users can test Voxtral for free by downloading the API from Hugging Face or trying the models in Mistral's chatbot Le Chat. API integration for applications starts at $0.001 per minute, according to the company.
This release follows Mistral's announcement of Magistral last month, its first family of reasoning models that solve problems step-by-step for enhanced reliability.
Mistral, a leading AI firm in Europe, is renowned for its advocacy of open-source AI models. Earlier this month, TechCrunch reported that the company is negotiating to raise up to $1 billion in equity from investors such as Abu Dhabi's MGX fund.
Ollie bets privacy focus to win AI assistant race
To be genuinely helpful, an AI assistant must understand its user deeply. Ollie, a personal assistant designed for daily life, operates on the premise that this doesn’t require surrendering your data or compromising your privacy.While certain enterpr
How AI LIVE: London Will Explore AI & Industrial Automation
The summit will convene C-suite executives from around the globe to address pressing challenges in global industries, ranging from AI-driven disruption to economic volatility.AI LIVE: The London Summit will gather over 2,000 international leaders und





Home






