
French AI lab Mistral AI unveils its new multimodal model, Mistral Large 4, with plans for an open-weight release following safety testing.
AI-generated summary
Mistral AI is a French artificial intelligence company that recently reached a €21 billion valuation. The company focuses on developing large language models as an alternative to US and Chinese counterparts.
The race between open and closed AI is on, and Europe is still in the mix. On Tuesday, French AI lab Mistral AI has released Mistral Large 4, a new large multimodal model aiming to leapfrog both American and Chinese rivals — following what French president Macron has described as “a third way in AI.”
Amid a growing divide between closed models that can be unplugged and open models that are often made in China, Mistral is positioning ML4 as an alternative to both. Nicknamed Le Chonk in reference to its one trillion parameters, ML4 is definitely not small; but is not an open-weight model yet. For the time being, it can only be accessed via a public guardrail endpoint, but Mistral plans to make its weights available in just three weeks, after safety testing is complete.
“In the meantime, we’ll work with trusted partners and governments to make sure that the open-source weights can be used to defend, but not to [perform] malicious attacks,” Mistral VP Science Pierre Stock told TechCrunch.
Security concerns have been mounting in recent months, particularly among Mistral’s core audience — enterprises and institutions. At the same time, an open-weight model is easier to audit, Stock said.
Another important behind-the-scenes aspect is that ML4 was trained entirely on Mistral’s compute; using only 4,000 NVIDIA GPUs “which is two to three times less than our Chinese competitors, and significantly less than the closed source competitors,” Stock said.
With benchmark results still pending, Mistral hopes ML4 will be best in class among open-weight models, especially outside of China, but not only, Stock said. Thanks to focused training, it could also outperform closed models in specific areas that are key to its customers, and where multimodal capabilities can add value.
According to Stock, ML4’s optimized use cases include cybersecurity and finance, but also chip design, which is core to two of Mistral’s main backers — Dutch giant ASML, which led its Series C, and Samsung, which led its Series D last month at a €21 billion valuation (about $24.39 billion).
AI outlook — possibilities, not facts
Release of open weights for Mistral Large 4.
Very likely · Within weeks

Starting October 9th, Google is restructuring its Gemini subscription plans. Free users will be limited to the Flash Lite model, while the $4.99/month AI Plus tier loses Gemini Pro access. Advanced models now require higher-priced Pro or Ultra subscriptions.

Meta is testing a new full-screen, video-first Facebook interface in India that launches directly into Reels. The test aims to increase engagement with short-form video, allowing users to opt out and return to the traditional feed if preferred.
Google may expand its 'Call for Me' AI feature beyond business calls to allow users to send personal messages via automated calls, such as telling a friend they will be late or asking about dinner plans, based on an APK teardown by Android Authority that revealed a 'Gemini Calling' screen and granular app permissions for controlling Gemini's access to phone functions.

OpenAI will roll out an invisible text watermark for ChatGPT and Codex in the European Union to comply with the EU AI Act's transparency rules, which took effect on August 2. The watermark, called textGrain, shapes word choices to create a detectable pattern without affecting model performance. It will be available to eligible users in the EU over the coming weeks and to API developers worldwide starting today, though off by default. OpenAI notes the watermark can be reduced by editing and does not prove human authorship if missing.

AI chip startup Etched is reviewing acquisition offers ranging from $40 billion to $50 billion just months after raising $700 million at a $21 billion valuation, with investors attracted by its proprietary hardware systems, Nvidia-challenging technology, and talent poaching from the chip giant.

OpenAI has announced solutions to major mathematical problems including the Navier-Stokes Millennium Prize Problem, but its rapid announcements and lack of transparency have sparked controversy among mathematicians who accuse the company of scooping, spying, and violating academic norms, despite forming an advisory group to improve community relations.