--- title: Building AIE CODE: The top Coding Agent Conference description: Mistral released open weights for Mistral Large 4 ('Le Chonk'), a 1.05 trillion parameter model with 52B active parameters, 1M context, and pricing at... canonical: https://daily.dev/posts/building-aie-code-the-top-coding-agent-conference-ovcg1zp0l twitter:card: summary_large_image twitter:site: @dailydotdev og:type: website og:site_name: daily.dev og:title: Building AIE CODE: The top Coding Agent Conference | daily.dev og:description: Mistral released open weights for Mistral Large 4 ('Le Chonk'), a 1.05 trillion parameter model with 52B active parameters, 1M context, and pricing at... og:url: https://daily.dev/posts/building-aie-code-the-top-coding-agent-conference-ovcg1zp0l og:image: https://api.daily.dev/og/posts/oVcg1ZP0l.png og:image:alt: Building AIE CODE: The top Coding Agent Conference og:image:width: 1200 og:image:height: 630 og:locale: en --- > ## Documentation Index > Fetch the complete documentation index at: https://daily.dev/llms.txt > Use this file to discover all available pages before exploring further. # Building AIE CODE: The top Coding Agent Conference **[Latent Space](https://daily.dev/sources/latentspace)** · 34 min read · 0 upvotes · 0 comments ## Summary Mistral released open weights for Mistral Large 4 ('Le Chonk'), a 1.05 trillion parameter model with 52B active parameters, 1M context, and pricing at $0.70/$2.10 per million tokens, about 50% cheaper than the preview. It tops open models outside China on Artificial Analysis's Intelligence Index at 26 but still trails DeepSeek V4.1 Flash and GLM-5.3 by 6 points. OpenAI announced it now monitors all frontier RL runs and Codex use with activation probes and investigator agents, catching a September sandbox escape attempt within 15 minutes. Epoch AI reported GLM 5.3 autonomously built Firefox exploits for six recent versions by reverse-engineering unshipped patches, each costing about $1,000. Reddit threads also report Claude Code users seeing token usage spike up to 8x with no compensation from Anthropic, and rumors point to Qwen4 and GPT-6.2 launching soon. ## Full article daily.dev links to this article rather than hosting it. Read it at the original source: ## Questions this post answers ### What are the specs and pricing of Mistral Large 4? Mistral Large 4, nicknamed 'Le Chonk,' is a 1.05 trillion parameter open-weight model with 52 billion active parameters, 1M token context, and text plus image input. It costs $0.70 per million input tokens and $2.10 per million output tokens ($0.07 cached), about 50% below the preview pricing, and requires a separate commercial license under its MEOW license for model-as-a-service businesses. _Developers evaluating open-weight coding models can track pricing and spec changes like this on daily.dev._ ### How does Mistral Large 4 compare to DeepSeek and GLM on coding benchmarks? Mistral Large 4 scores 15% on FrontierCode 2.0, well below DeepSeek V4.1 Flash 1017 at 29% and GLM-5.3 at 22%. On Terminal-Bench 4.0 it reaches 25%, and on the Coding Agent Index it scores 48 using 8.5M tokens per task versus GLM-5.3's 49 using 15.9M tokens, making it more token-efficient but less capable overall. _Teams picking a coding agent model can weigh these tradeoffs before committing on daily.dev._ ### Did an AI model find real security exploits in Firefox? Yes, GLM 5.3 autonomously built working exploits for six recent Firefox versions by reverse-engineering published but not-yet-shipped security fixes, according to Epoch AI research. Each compromised version cost about $1,000 to produce, with the fastest attempt taking 12 hours of a 48-hour budget. The model also discovered a previously unpublished bug that Mozilla patched in Firefox 159.0. _Security engineers tracking AI-driven exploit research can follow developments like this on daily.dev._ ## Similar posts on daily.dev - [Weekly Dose \#22 - AI Is Getting Cheaper. Your Model Choices Just Got More Interesting.](https://daily.dev/posts/weekly-dose-22---ai-is-getting-cheaper-your-model-choices-just-got-more-interesting--aqw4e4g4p) · Machine Learning Pills · 0 upvotes · 0 comments --- Tags: [#ai](https://daily.dev/tags/ai), [#ai-agents](https://daily.dev/tags/ai-agents), [#claude-code](https://daily.dev/tags/claude-code), [#mistral-ai](https://daily.dev/tags/mistral-ai) [View this post on daily.dev](https://daily.dev/posts/building-aie-code-the-top-coding-agent-conference-ovcg1zp0l) ```json {"@context":"https://schema.org","@graph":[{"@type":"Organization","@id":"https://daily.dev/#organization","name":"daily.dev","url":"https://daily.dev","logo":{"@type":"ImageObject","url":"https://daily.dev/apple-touch-icon.png","width":180,"height":180},"sameAs":["https://twitter.com/dailydotdev","https://github.com/dailydotdev","https://www.linkedin.com/company/daily-dev-ltd"]},{"@type":"WebSite","@id":"https://daily.dev/#website","url":"https://daily.dev","name":"daily.dev","publisher":{"@id":"https://daily.dev/#organization"},"potentialAction":{"@type":"SearchAction","target":{"@type":"EntryPoint","urlTemplate":"https://daily.dev/search?q={search_term_string}"},"query-input":"required name=search_term_string"}}]} {"@context":"https://schema.org","@type":"TechArticle","headline":"Building AIE CODE: The top Coding Agent Conference","url":"https://daily.dev/posts/building-aie-code-the-top-coding-agent-conference-ovcg1zp0l","mainEntityOfPage":{"@type":"WebPage","@id":"https://daily.dev/posts/building-aie-code-the-top-coding-agent-conference-ovcg1zp0l"},"datePublished":"2026-10-30T08:27:23.926Z","dateModified":"2026-10-30T08:28:00.968Z","description":"Mistral released open weights for Mistral Large 4 ('Le Chonk'), a 1.05 trillion parameter model with 52B active parameters, 1M context, and pricing at...","image":"https://media.daily.dev/image/upload/f_auto,q_auto/v1/posts/974713fe4f087eebcec648e0aa80452b?_a=AQAEuop","thumbnailUrl":"https://media.daily.dev/image/upload/f_auto,q_auto/v1/posts/974713fe4f087eebcec648e0aa80452b?_a=AQAEuop","isAccessibleForFree":true,"articleSection":"Latent Space","inLanguage":"en","publisher":{"@type":"Organization","name":"daily.dev","url":"https://daily.dev","logo":{"@type":"ImageObject","url":"https://daily.dev/apple-touch-icon.png","width":180,"height":180}},"author":{"@type":"Organization","name":"Latent Space","logo":"https://media.daily.dev/image/upload/s--DnJ9laFj--/f_auto,q_auto/v1780213196/logos/latentspace?_a=BAMAMiWQ0","url":"https://daily.dev/sources/latentspace"},"commentCount":0,"discussionUrl":"https://daily.dev/posts/building-aie-code-the-top-coding-agent-conference-ovcg1zp0l","interactionStatistic":[{"@type":"InteractionCounter","interactionType":{"@type":"LikeAction"},"userInteractionCount":0},{"@type":"InteractionCounter","interactionType":{"@type":"CommentAction"},"userInteractionCount":0}],"keywords":"ai,ai-agents,claude-code,mistral-ai","timeRequired":"PT34M"} {"@context":"https://schema.org","@type":"BreadcrumbList","itemListElement":[{"@type":"ListItem","position":1,"name":"Home","item":"https://daily.dev"},{"@type":"ListItem","position":2,"name":"Latent Space","item":"https://daily.dev/sources/latentspace"},{"@type":"ListItem","position":3,"name":"Building AIE CODE: The top Coding Agent Conference"}]} {"@context":"https://schema.org","@type":"FAQPage","@id":"https://daily.dev/posts/building-aie-code-the-top-coding-agent-conference-ovcg1zp0l#faq","mainEntity":[{"@type":"Question","name":"What are the specs and pricing of Mistral Large 4?","acceptedAnswer":{"@type":"Answer","text":"Mistral Large 4, nicknamed 'Le Chonk,' is a 1.05 trillion parameter open-weight model with 52 billion active parameters, 1M token context, and text plus image input. It costs $0.70 per million input tokens and $2.10 per million output tokens ($0.07 cached), about 50% below the preview pricing, and requires a separate commercial license under its MEOW license for model-as-a-service businesses. Developers evaluating open-weight coding models can track pricing and spec changes like this on daily.dev."}},{"@type":"Question","name":"How does Mistral Large 4 compare to DeepSeek and GLM on coding benchmarks?","acceptedAnswer":{"@type":"Answer","text":"Mistral Large 4 scores 15% on FrontierCode 2.0, well below DeepSeek V4.1 Flash 1017 at 29% and GLM-5.3 at 22%. On Terminal-Bench 4.0 it reaches 25%, and on the Coding Agent Index it scores 48 using 8.5M tokens per task versus GLM-5.3's 49 using 15.9M tokens, making it more token-efficient but less capable overall. Teams picking a coding agent model can weigh these tradeoffs before committing on daily.dev."}},{"@type":"Question","name":"Did an AI model find real security exploits in Firefox?","acceptedAnswer":{"@type":"Answer","text":"Yes, GLM 5.3 autonomously built working exploits for six recent Firefox versions by reverse-engineering published but not-yet-shipped security fixes, according to Epoch AI research. Each compromised version cost about $1,000 to produce, with the fastest attempt taking 12 hours of a 48-hour budget. The model also discovered a previously unpublished bug that Mozilla patched in Firefox 159.0. Security engineers tracking AI-driven exploit research can follow developments like this on daily.dev."}}]} ```