--- title: "Mistral publishes weights, halves prices, and makes resellers sign a deal; seven Chinese models outrank it" description: "Le Chonk's open weights arrive with a resale license, a memory wall and a scoreboard that still ranks seven Chinese models ahead. Sovereign buyers get AI no provider can switch off; Mistral gets a funnel into its own paid serving." publisher: "The Inference" section: "Model" published: 2026-10-30T14:48:28.006Z modified: 2026-10-30T14:48:28.006Z canonical: https://theinference.org/article/mistral-publishes-weights-halves-prices-and-makes-resellers-sign-a-deal-seven-chinese-models-outrank-it language: en keywords: "Mistral, Open Weights, AI Benchmarks, Model Pricing, Sovereign AI" --- # Mistral publishes weights, halves prices, and makes resellers sign a deal; seven Chinese models outrank it > Le Chonk's open weights arrive with a resale license, a memory wall and a scoreboard that still ranks seven Chinese models ahead. Sovereign buyers get AI no provider can switch off; Mistral gets a funnel into its own paid serving. On October 29 [Mistral](https://theinference.org/markets/companies/mistral-ai), the Paris-based AI lab, [published the weights of Mistral Large 4](https://mistral.ai/news/mistral-large-4-oss/), the trillion-parameter model it calls Le Chonk, and cut its API price roughly in half the same day [1][2]. Read the license alongside the generosity. Large 3 shipped under Apache 2.0; Large 4 ships under a modified MIT license that is free to run, fine-tune and commercialize on one condition: any business offering the model to third parties as a service [must sign a separate commercial agreement with Mistral](https://winbuzzer.com/2026/10/29/mistral-releases-le-chonk-ai-for-self-hosted-cyber-defense-xcxwbn/) [3]. Self-hosting has its own floor. The compressed 2-bit build still needs 512GB of unified memory, a Mac Studio M5 Ultra that [costs $22,299 in the US](https://www.martincid.com/technology-sv/mistral-large-4-le-chonk-open-weights-mac-studio/) [5]. So the weights are free to evaluate, and the money changes hands when a customer scales: rent from Mistral, or pay the toll to resell. *Chart: **The v5 index puts the closed frontier 32 points above Mistral Large 4** Bar chart ranking AI models on the Artificial Analysis v5 Intelligence Index from 58 for Claude Opus 5.5 down to 25 for Meta's Muse Spark Open, with Mistral Large 4 highlighted at 26.* *Artificial Analysis Intelligence Index v5, composite points, higher is better. Kimi K3.1 Pro's weights are unreleased until November 4. Sources: Trending Topics; TechTimes. [2][4]* ## The scoreboard is the risk What the toll buys sits behind rivals. On the Cyber Index from Artificial Analysis, an independent benchmarking firm, which measures whether a model can find, reproduce and patch code vulnerabilities, Mistral trails both [DeepSeek](https://theinference.org/markets/companies/deepseek) and [Xiaomi](https://theinference.org/markets/companies/xiaomi)'s MiMo [3]. On [the redesigned v5 Intelligence Index](https://www.techtimes.com/articles/329670/20261029/agentic-ai-benchmarks-widen-proprietary-model-edge-12-points-artificial-analysis-v5-overhaul.htm) it scores 26, [eighth among open models, behind seven Chinese models and just ahead of Meta](https://www.trendingtopics.eu/mistral-large-4-open-weights-artificial-analysis-meta-china-ranking/), against a closed frontier led by [Anthropic](https://theinference.org/markets/companies/anthropic)'s Claude Opus 5.5 and [OpenAI](https://theinference.org/markets/companies/openai)'s GPT-6 Astra [2][4]. Our earlier piece this week, "Mistral now speaks OpenAI's agent format; its flagship scores 25 against OpenAI's 57", caught the preview at 25; the released model scores 26, one point more, and its position has not changed since that first test [2]. The v5 overhaul widened the gap between the best closed and the best open model from 3 points to 12, but that gap counts Kimi K3.1 Pro's 46, and Kimi is scheduled to release its weights on November 4 [4]. Against the open models anyone can download today, DeepSeek V4.1 Flash and GLM-5.3 at 32, the released field sits 26 points below the leader [4]. The open leaderboard resets this week. *Chart: **Mistral's cyber defense claim trails both Chinese open rivals** Bar chart of Artificial Analysis Cyber Index scores: DeepSeek V4.1 Flash 57, Xiaomi MiMo-V2.6-Pro 56, Mistral Large 4 highlighted at 51.* *Artificial Analysis Cyber Index, points, higher is better; the index measures whether a model can find, reproduce and patch code vulnerabilities. Source: WinBuzzer. [3]* ## Who the money touches The buyers arrived anyway. Orange Cyberdefense is [integrating a cyber-focused variant into selected services](https://techstartups.com/2026/10/29/mistral-ai-releases-large-4-a-1-trillion-parameter-open-weight-model-with-a-cybersecurity-focus/) [8]. La Banque Postale, a French citizen bank, calls technological sovereignty a prerequisite for customer trust [1]. [Microsoft](https://theinference.org/markets/companies/microsoft) Foundry is the first cloud to serve the model [1]. Listed exposure runs wider: [Nvidia](https://theinference.org/markets/companies/nvidia), whose [3,800 Grace Blackwell GPUs trained the model](https://siliconangle.com/2026/10/29/mistral-releases-open-weights-for-mistral-large-4-providing-sovereign-control-for-the-companys-le-chonk-ai-model/) [6]; [Apple](https://theinference.org/markets/companies/apple), which sells the machine the memory wall points at [5]; Orange; and Xiaomi, which beats Mistral on the cyber score [3]. [Samsung led Mistral's September round](https://www.cnbc.com/2026/09/08/mistral-ai-funding-valuation-samsung.html) [9]. ## The toll is worth what the road is worth Arthur Mensch's pitch is the capital: "Other labs have raised hundreds of billions to develop equivalent capabilities; [in three years, Mistral raised roughly $6 billion](https://app.dealroom.co/news/note/sovereign-ai-gpu-fleets-and-an-ipo-in-the-next-few-years-mensch-on-mistral-s-vertical-integration-bet)", with an IPO within a few years [6][7]. The €3 billion September round at a €21 billion valuation is what that discipline has bought so far [9]. The giveaway is the marketing, the license toll and the halved API are the margin, and the scoreboard is the risk. Sovereign demand is real: institutions want AI no provider can switch off, and the arrival of Orange Cyberdefense, La Banque Postale and Microsoft alongside a model that trails rivals suggests demand outweighs capability in this market. Mistral is doing the work in between, tailoring the weights to a bank's detection logic and a defender's threat environment through its Forge and Studio products [1], and that work is where sovereign value gets created. But a toll prices the road, and this week the road ranks eighth among open weights, behind labs that answer to a different government. Watch November 4: if Kimi's scheduled weights close the gap Mistral claims to lead, the toll buys less. ## Takeaway Mistral's open release pairs a resale toll and halved API prices with real sovereign demand, but the model still ranks eighth among open weights; whether Kimi's November 4 weights close the gap decides what the toll is worth. ## Sources 1. [Mistral AI, Mistral Large 4 (Le Chonk): Built for Coders, Trusted by Defenders, 29 October 2026](https://mistral.ai/news/mistral-large-4-oss/) 2. [Trending Topics, Mistral Large 4 Goes Open Weight, Lands Ahead of Meta but Behind China, October 2026](https://www.trendingtopics.eu/mistral-large-4-open-weights-artificial-analysis-meta-china-ranking/) 3. [WinBuzzer, Mistral Releases Mistral Large 4 for Self-Hosted Cyber Defense, 29 October 2026](https://winbuzzer.com/2026/10/29/mistral-releases-le-chonk-ai-for-self-hosted-cyber-defense-xcxwbn/) 4. [TechTimes, Agentic AI Benchmarks Widen Proprietary-Model Edge to 12 Points in Artificial Analysis v5 Overhaul, 29 October 2026](https://www.techtimes.com/articles/329670/20261029/agentic-ai-benchmarks-widen-proprietary-model-edge-12-points-artificial-analysis-v5-overhaul.htm) 5. [Martin Cid Magazine, Mistral Large 4 is free to download, and it fits on one $22,299 Mac, 29 October 2026](https://www.martincid.com/technology-sv/mistral-large-4-le-chonk-open-weights-mac-studio/) 6. [SiliconANGLE, Mistral releases open weights for Mistral Large 4, providing sovereign control for the company's 'Le Chonk' AI model, 29 October 2026](https://siliconangle.com/2026/10/29/mistral-releases-open-weights-for-mistral-large-4-providing-sovereign-control-for-the-companys-le-chonk-ai-model/) 7. [Dealroom, Sovereign AI, GPU fleets and an IPO in the next few years: Mensch on Mistral's vertical-integration bet, 28 October 2026](https://app.dealroom.co/news/note/sovereign-ai-gpu-fleets-and-an-ipo-in-the-next-few-years-mensch-on-mistral-s-vertical-integration-bet) 8. [TechStartups (Crunchbase), Orange Cyberdefense partnership announcement, 29 October 2026](https://techstartups.com/2026/10/29/mistral-ai-releases-large-4-a-1-trillion-parameter-open-weight-model-with-a-cybersecurity-focus/) 9. [CNBC, Mistral bags $24 billion valuation as Samsung leads funding for Europe's AI champion, 8 September 2026](https://www.cnbc.com/2026/09/08/mistral-ai-funding-valuation-samsung.html) --- Mistral publishes weights, halves prices, and makes resellers sign a deal; seven Chinese models outrank it — The Inference. Canonical: https://theinference.org/article/mistral-publishes-weights-halves-prices-and-makes-resellers-sign-a-deal-seven-chinese-models-outrank-it