AI··1 min readTechnical readSource-backed

Mistral says "Le Chonk" can challenge the best AI models

French company Mistral has released a new freely available model named Mistral Large 4, nicknamed Le Chonk, which the company claims can challenge top models from the US and China.

By
mistral header 2 1152x648 1
Ars Technica
WIREUNWIRED BRIEF

French company Mistral has released a new freely available model named Mistral Large 4, nicknamed Le Chonk, which the company claims can challenge top models from the US and China. The new one trillion-parameter model is currently in preview, with a final version expected by the end of the month.

While built as a general-purpose model, Mistral optimized it for coding, cyberdefense, manufacturing, finance, and electrical engineering. Mistral states that the model was trained from scratch, positioning it as the most capable open-weight model outside China. Mistral cofounder and chief scientist Guillaume Lample told WIRED that the company aims to improve models in specialized domains where other labs spend less focus.

Because open-weight models only cost as much as the compute they consume, they are cheaper for businesses to run than closed alternatives. Mistral believes the release eliminates the remaining reasons for businesses to hesitate on open source software. The release follows a period of financial growth for Mistral, which recently raised a $3.3 billion funding round at a $24 billion valuation.

While the company has historically lagged behind US labs in performance and release frequency due to having less capital, it generates revenue through cloud pay-as-you-go fees and custom

STORY JOURNEY

The story so far

Earlier coverage connected by shared names and topics. Follow the dates to see how coverage developed; these links do not imply every article describes the same event.

  1. A new study from Graphite reveals that artificial intelligence models continue to use distinct writing habits.

    A new study from the marketing firm Graphite reveals that artificial intelligence models still use distinct writing habits, despite dropping older clues…

    Connected by: WhileEARLIER COVERAGE
  2. GRC 2026 C 02774large
    NASA Conducts Tunnel Tests at Glenn Research Center to Study Supercooled Large Droplet Icing Hazards

    NASA is testing methods to improve aviation safety by studying a rare icing hazard called supercooled large droplet icing. This hazard happens…

    Connected by: LargeEARLIER COVERAGE
  3. Editorial illustration for French AI lab Mistral AI releases Mistral Large 4, a one-trillion-parameter multimodal model trained on 4,000 GPUs.
    French AI lab Mistral AI releases Mistral Large 4, a one-trillion-parameter multimodal model trained on 4,000 GPUs.

    French artificial intelligence lab Mistral AI released a new large multimodal model called Mistral Large 4 on Tuesday. Nicknamed Le Chonk, the…

    Connected by: Mistral LargeEARLIER COVERAGE
  4. chatgpt icon 1152x648 1789154124
    OpenAI will automatically watermark ChatGPT text outputs within the European Union to comply with the EU AI Act.

    OpenAI announced it will automatically watermark text generated by ChatGPT within the European Union. This step helps the company comply with the…

    Connected by: BecauseEARLIER COVERAGE
  5. NOW
    Mistral says "Le Chonk" can challenge the best AI modelsCURRENT UPDATE
A WireUnwired summary is already available. You can still write your own brief below for the reader community.
SOURCE & ATTRIBUTION
Ars Technica

Original author: Joel Khalili, wired.com

Original publication: 7 October 2026 19:42

Read full story on Ars Technica ↗
RELATED STORIES
Editorial illustration for Millions of Hype, Four Uber Rides: Why AI Agents Are StallingTechCrunchMillions of Hype, Four Uber Rides: Why AI Agents Are StallingEditorial illustration for Titan OS Secures Funding to Challenge Tech Giants in the Living RoomTechCrunchTitan OS Secures Funding to Challenge Tech Giants in the Living RoomEditorial illustration for Anthropic Is Building a Cryptographic Guard Against Rogue AI Model SwappingTechCrunchAnthropic Is Building a Cryptographic Guard Against Rogue AI Model Swapping
Original sourceArs Technica ↗