Mistral AI introduced two new transcription models—Voxtral Mini Transcribe 2 and Voxtral Realtime—designed to run on edge devices such as phones, laptops, and wearables. The compact models prioritize privacy by keeping data local, and they deliver low‑latency performance, with the realtime model achieving less than 200 milliseconds of delay. Available via Mistral’s API and on Hugging Face, the models support 13 languages and can be customized for specific vocabularies, offering accuracy comparable to larger systems while maintaining speed and user control. Leia mais →

Paris‑based Mistral AI announced two new speech‑to‑text models, Voxtral Mini Transcribe V2 and Voxtral Realtime, that operate with just four billion parameters. The models can run on a phone or laptop, support real‑time transcription within 200 milliseconds, and translate across 13 languages. By keeping the models small, Mistral aims to lower costs, improve privacy, and offer an open‑source alternative to larger U.S. offerings. The rollout reflects the company’s strategy of delivering specialized, efficient AI tools while emphasizing European sovereignty in the fast‑moving generative‑AI market. Leia mais →