Voxtral Voxtral introduces advanced speech understanding models, available in 24B and 3B variants, under the Apache 2.0 license and via an API, aiming to make voice interaction more natural and capable. These models surpass traditional limitations by offering superior transcription accuracy, native semantic understanding, multilingual support, and function-calling capabilities, bridging the gap between costly proprietary APIs and less capable open-source solutions. Voxtral also retains strong text understanding capabilities and offers enterprise-focused features for advanced deployment and customization.

• Voxtral models (24B and 3B variants) offer state-of-the-art speech understanding.
• They are released under the Apache 2.0 license and available via API.
• Voxtral provides high accuracy and semantic understanding at a lower cost than comparable proprietary APIs.
• Key capabilities include long-form context handling, built-in Q&A and summarization, multilingual support, function-calling, and strong text understanding.
• Voxtral outperforms existing models like Whisper, GPT-4o mini, and Gemini 2.5 Flash on various benchmarks.
• The models can be downloaded locally, accessed via API, or tested on Le Chat.
• Enterprise features include private deployment, domain-specific fine-tuning, advanced context options, and dedicated integration support.
• Future updates will include speaker segmentation, audio markups, word-level timestamps, and non-speech audio recognition. Continue reading https://foxvector.com/articles/4228572e-45be-4106-ae70-3e5d33da9170