Mistral AI just released Mistral OCR, a dedicated API built to parse messy documents, math equations, and tables into structured markdown. It can process up to 2,000 pages per minute at $1 per 1,000 pages, and is already powering document understanding inside Le Chat.
Mistral just rolled out dedicated iOS and Android apps for its le Chat assistant alongside a major platform upgrade. It packs web search, document processing, Flux Ultra image generation, and a new Flash Answers mode that replies over 10x faster than ChatGPT.
Mistral cofounder and CEO Arthur Mensch just shut down buyout rumors at Davos, stating the European AI champion is firmly "not for sale" and plans to pursue an IPO instead. Amid industry chatter about startup consolidation, the company is also opening a Singapore office and continuing its aggressive expansion across the US and Europe.
Mistral AI just rolled out a major beta update to le Chat, packing web search with citations, a collaborative Canvas workspace, and Flux Pro image generation into its free assistant. Powered by the new Pixtral Large multimodal model for complex document and image reasoning, it also lets users build workflow agents—giving ChatGPT a serious free run for its money.
Mistral AI just dropped Mistral Large 2, a 123B-parameter flagship model built with a 128k context window designed to run efficiently on a single node. It delivers major gains in coding and math that rival GPT-4o and Claude 3 Opus, while actively curbing hallucinations. Non-commercial weights are free to download, offering serious frontier power without giant infrastructure.
Apple's research team just released open DCLM models with 7B and 1.4B parameters on Hugging Face, sharing weights, code, and training datasets. The 7B model scored 63.7% on MMLU, outperforming Mistral-7B while using 40% less compute. It shows Apple's data curation prowess, though it's research-focused rather than on-device software.
Mistral AI just launched Codestral Mamba, a code-focused model built with Albert Gu and Tri Dao that swaps Transformers for the Mamba architecture. It delivers linear-time inference and handles tested in-context retrieval up to 256k tokens without slowdowns. Even better, Mistral open-sourced it under Apache 2.0 for free local deployment.
Mistral AI just dropped Codestral, its first-ever generative model fine-tuned specifically for coding, sporting 22B parameters and a 32k context window. Fluent in over 80 programming languages, it shines at fill-in-the-middle completions and repo-level tasks. Why it matters: Developers get an open-weight, low-latency engine directly inside their IDEs, though commercial deployments require a custom license.
Hugging Face just released Idefics2, an open-source 8B vision-language model licensed under Apache 2.0. It preserves native image resolutions, significantly upgrades OCR, and punches way above its weight class on visual QA benchmarks against 30B+ rivals. Why it matters: Developers get an efficient multimodal base model ready to fine-tune directly in Transformers.
Mistral AI just dropped its largest open-source model yet, releasing Mixtral 8x22B via an unceremonious torrent link on X. The sparse MoE model features 176 billion total parameters with 44 billion active per token and a 65,000-token context window. Released under a permissive Apache 2.0 license, it delivers massive compute efficiency to open-source builders without corporate guardrails.
Microsoft just announced a multi-year partnership with French startup Mistral AI, making its new flagship model Mistral Large available first on Azure AI. Mistral gets Azure supercomputing power for model training, while Azure customers can run the model via MaaS alongside OpenAI's lineup. Why it matters: Microsoft is aggressively diversifying its cloud AI portfolio beyond OpenAI.
Mistral CEO Arthur Mensch just confirmed that the mystery open-source model “miqu-1-70b”—which scored near GPT-4 on benchmarks after surfacing on 4chan and Hugging Face—is indeed an old Mistral checkpoint retrained from Llama 2, leaked by an early customer’s employee. Why it matters: it proves open-source AI is closing the frontier performance gap much faster than expected.
Brave just upgraded its built-in browser assistant Leo, making Mistral's open-source Mixtral 8x7B the default model for all users. The upgrade brings sharper coding, larger context handling, and fewer hallucinations, all while maintaining Brave's strict privacy setup that discards chats immediately without requiring a login. Free users get access with rate limits, while Premium subscribers get higher usage caps.
French startup Mistral AI just raised a $113 million seed round at a $260 million valuation led by Lightspeed Venture Partners, only four weeks after launching. Founded by former DeepMind and Meta researchers, the team plans to build open-source generative models using publicly available data starting in 2024. Why it matters: Mistral wants to prove open-source infrastructure can challenge proprietary giants in the enterprise market.
