Paris-based Mistral releases "Le Chonk," a strong coding model that still trails China overall
Source: The Decoder (opens in a new tab) · Maximilian Schreiner
Intel Summary
Mistral has released the open weights for Large 4, an open-weight, trillion-parameter model trained entirely in Europe nicknamed "Le Chonk." Mistral claims the model outperforms open competitors at coding tasks, while independent testing indicates it operates quickly and at relatively low cost. However, it trails leading Chinese models, posting a score of 26 compared to 32 from DeepSeek V4.1 Flash and GLM-5.3.
Why It Matters
The release provides developers with an independently trained, large-scale European alternative for code generation and open-weight deployments. The performance gap behind leading Chinese models highlights continued competitive pressure in frontier open-weight foundation models.
Organizations & Entities
- China
- DeepSeek-V4.1-Flash
- Europe
- GLM-5.3
- Mistral
Topics
Related Intelligence
- ReportAlso involving Mistral
Mistral Says Its AI Model ‘Le Chonk’ Has Leapfrogged China in the Open-Weight Race
Mistral has announced a new open-weight AI model named 'Le Chonk'. According to claims by the French AI laboratory reported by WIRED, the model matches or outperforms top open-weight models from Chinese developers DeepSeek and Z.ai specifically in coding and cybersecurity tasks.
WIRED - ReportAlso involving China
OpenAI’s Astra Model Becomes Top Distillation Target for China’s AI Labs
Chinese artificial intelligence developers, including DeepSeek and Moonshot AI, are relying heavily on outputs from OpenAI's GPT-6 Astra to train their own latest models through distillation, according to reporting from The Information. The practice remains a primary driver of capability gains for Chinese open-weight models despite persistent industry criticism.
The Information - ReportAlso involving Europe
Le Chonk: Built for Coders, Trusted by Defenders.
Mistral AI has announced the open-weight release of Mistral Large 4, codenamed Le Chonk. The company states that the European-trained release is a one-trillion-parameter multimodal model optimized for coding tasks.
Mistral AI - DevelopmentNewAlso involving China
DeepSeek collaborates with Huawei to design AI chips
DeepSeek is collaborating closely with Huawei Technologies to guide the architectural design of future AI semiconductors, according to reporting from The Information. Huawei engineers are reportedly incorporating DeepSeek's technical input on calculation and inter-processor data transfer into AI chips planned for release next year and beyond, aiming to support frontier AI development independently of Nvidia hardware. Claims are as reported; this summary makes no determination about accuracy or significance.