Mistral says “Le Chonk” can challenge the best AI models

AI-rewritten: This is a summary of an article from Ars Technica, rewritten by AI (Qwen, running locally) to make it easier to read. The facts come from the original article – read it for the full story.

Ars Technica •
Joel Khalili, wired.com
• October 7, 2026

French company Mistral has released "Le Chonk," a new 1 trillion-parameter AI model designed to compete with top models from the United States and China. Guillaume Lample, cofounder and chief scientist at Mistral, told WIRED that the model is optimized for coding, cyberdefense, and specific industries like manufacturing and finance. He noted that other labs often do not focus on these many domains. The company claims Le Chonk was trained from scratch, unlike some Chinese models accused by the US government of using distillation to close performance gaps with proprietary systems.

Mistral states that its open-weight model is currently cheaper for businesses to run because costs depend only on compute consumption. Andrea Renda, director of research at the Centre for European Policy Studies, told WIRED in July that the EU’s strategy for technological sovereignty and US hostility create a favorable position for Mistral. The French lab recently raised $3.3 billion at a $24 billion valuation, with earnings reportedly increasing 20-fold in the last year.

Tensions over AI access have risen following temporary restrictions placed by the Trump administration on models from OpenAI and Anthropic due to concerns about cyberattacks. Mistral argues that relying on closed proprietary models risks sudden loss of access, regardless of location. Lample emphasized that owning the model matters even for US companies, as using a closed system offers no guarantee it will remain available tomorrow.

Source: Ars Technica •
Joel Khalili, wired.com
• October 7, 2026

Read the original article at Ars Technica →

Leave Comment

Your email address will not be published. Required fields are marked *