--- title: Claude Fable 5.2 is Absolutely Insane... | daily.dev description: Anthropic released Claude Fable 5.2 (with Mythos 5.2), a model the creator calls better than Opus 5.5 while costing twice as much per token but less per... canonical: https://daily.dev/posts/claude-fable-5-2-is-absolutely-insane--twgdlcnsz twitter:card: summary_large_image twitter:site: @dailydotdev og:type: website og:site_name: daily.dev og:title: Claude Fable 5.2 is Absolutely Insane... | daily.dev og:description: Anthropic released Claude Fable 5.2 (with Mythos 5.2), a model the creator calls better than Opus 5.5 while costing twice as much per token but less per... og:url: https://daily.dev/posts/claude-fable-5-2-is-absolutely-insane--twgdlcnsz og:image: https://api.daily.dev/og/posts/TwGDLcNSZ.png og:image:alt: Claude Fable 5.2 is Absolutely Insane... 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. # Claude Fable 5.2 is Absolutely Insane... **[NeuralNine](https://daily.dev/sources/neuralnine)** · 18 min read · 0 upvotes · 0 comments ## Summary Anthropic released Claude Fable 5.2 (with Mythos 5.2), a model the creator calls better than Opus 5.5 while costing twice as much per token but less per completed task. It generates output up to 30% faster than the previous Fable version. On the Artificial Analysis benchmark it scores highest, with Opus 5.5 third. In a release-notes example, Fable 5.2 with over 1,000 sub-agents running for 8 days built an entire operating system from scratch. In a weight-space programming test (manually writing neural network weights to play Breakout), Fable scored around 493, below GPT-6 Astra's 517 but above Opus 5.5's roughly 200-250. ## Full article daily.dev links to this article rather than hosting it. Read it at the original source: ## Questions this post answers ### How does Claude Fable 5.2 pricing compare to Opus 5.5? Fable 5.2 costs about twice as much as Opus 5.5 per token, but roughly half as much per completed task. At high effort, Fable 5.2 scores about level with Opus 5.5 at max effort while using fewer tokens overall, making it effectively cheaper on a per-task basis despite the higher token price. _Developers weighing Claude model costs for real workloads can track pricing shifts like this on daily.dev._ ### What is weight space programming and how did Claude Fable 5.2 perform at it? Weight space programming means an LLM manually writes or edits a neural network's weights directly, without training via gradient descent. Fable 5.2 was given a model architecture description for playing Breakout and hand-crafted the weights, achieving an average score of about 493, compared to roughly 200-250 for Opus 5.5 and 517 for GPT-6 Astra. _Anyone exploring novel LLM capabilities beyond code generation can follow experiments like this on daily.dev._ ### Is Claude Fable 5.2 the same model as Claude Mythos 5.2? Yes, Fable and Mythos 5.2 are the same underlying model, but Fable includes additional safeguards around biology and cybersecurity topics to prevent misuse for creating dangerous weapons, while Mythos lacks those extra restrictions. _Teams picking between Claude model variants for sensitive use cases can keep tabs on these distinctions via daily.dev._ ## Similar posts on daily.dev - [Anthropic’s Opus 5 is almost Fable 5](https://daily.dev/posts/anthropic-s-opus-5-is-almost-fable-5-vlen80a1c) · The New Stack · 9 upvotes · 1 comments --- Tags: [#ai](https://daily.dev/tags/ai), [#ai-coding](https://daily.dev/tags/ai-coding), [#claude](https://daily.dev/tags/claude), [#anthropic](https://daily.dev/tags/anthropic), [#gpt](https://daily.dev/tags/gpt) [View this post on daily.dev](https://daily.dev/posts/claude-fable-5-2-is-absolutely-insane--twgdlcnsz) ```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":"Claude Fable 5.2 is Absolutely Insane...","url":"https://daily.dev/posts/claude-fable-5-2-is-absolutely-insane--twgdlcnsz","mainEntityOfPage":{"@type":"WebPage","@id":"https://daily.dev/posts/claude-fable-5-2-is-absolutely-insane--twgdlcnsz"},"datePublished":"2026-11-10T22:41:02.836Z","dateModified":"2026-11-10T22:42:02.799Z","description":"Anthropic released Claude Fable 5.2 (with Mythos 5.2), a model the creator calls better than Opus 5.5 while costing twice as much per token but less per...","image":"https://i.ytimg.com/vi/W-kpXELz6DI/sddefault.jpg","thumbnailUrl":"https://i.ytimg.com/vi/W-kpXELz6DI/sddefault.jpg","isAccessibleForFree":true,"articleSection":"NeuralNine","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":"NeuralNine","logo":"https://media.daily.dev/image/upload/s--bMYO07OB--/f_auto,q_auto/v1777800757/logos/neuralnine?_a=BAMAMiWQ0","url":"https://daily.dev/sources/neuralnine"},"commentCount":0,"discussionUrl":"https://daily.dev/posts/claude-fable-5-2-is-absolutely-insane--twgdlcnsz","interactionStatistic":[{"@type":"InteractionCounter","interactionType":{"@type":"LikeAction"},"userInteractionCount":0},{"@type":"InteractionCounter","interactionType":{"@type":"CommentAction"},"userInteractionCount":0}],"keywords":"ai,ai-coding,claude,anthropic,gpt","timeRequired":"PT18M","video":{"@type":"VideoObject","name":"Claude Fable 5.2 is Absolutely Insane...","description":"Anthropic released Claude Fable 5.2 (with Mythos 5.2), a model the creator calls better than Opus 5.5 while costing twice as much per token but less per...","thumbnailUrl":"https://i.ytimg.com/vi/W-kpXELz6DI/sddefault.jpg","uploadDate":"2026-11-10T22:41:02.836Z","duration":"PT18M","url":"https://api.daily.dev/r/TwGDLcNSZ","embedUrl":"https://www.youtube.com/embed/W-kpXELz6DI"}} {"@context":"https://schema.org","@type":"BreadcrumbList","itemListElement":[{"@type":"ListItem","position":1,"name":"Home","item":"https://daily.dev"},{"@type":"ListItem","position":2,"name":"NeuralNine","item":"https://daily.dev/sources/neuralnine"},{"@type":"ListItem","position":3,"name":"Claude Fable 5.2 is Absolutely Insane..."}]} {"@context":"https://schema.org","@type":"FAQPage","@id":"https://daily.dev/posts/claude-fable-5-2-is-absolutely-insane--twgdlcnsz#faq","mainEntity":[{"@type":"Question","name":"How does Claude Fable 5.2 pricing compare to Opus 5.5?","acceptedAnswer":{"@type":"Answer","text":"Fable 5.2 costs about twice as much as Opus 5.5 per token, but roughly half as much per completed task. At high effort, Fable 5.2 scores about level with Opus 5.5 at max effort while using fewer tokens overall, making it effectively cheaper on a per-task basis despite the higher token price. Developers weighing Claude model costs for real workloads can track pricing shifts like this on daily.dev."}},{"@type":"Question","name":"What is weight space programming and how did Claude Fable 5.2 perform at it?","acceptedAnswer":{"@type":"Answer","text":"Weight space programming means an LLM manually writes or edits a neural network's weights directly, without training via gradient descent. Fable 5.2 was given a model architecture description for playing Breakout and hand-crafted the weights, achieving an average score of about 493, compared to roughly 200-250 for Opus 5.5 and 517 for GPT-6 Astra. Anyone exploring novel LLM capabilities beyond code generation can follow experiments like this on daily.dev."}},{"@type":"Question","name":"Is Claude Fable 5.2 the same model as Claude Mythos 5.2?","acceptedAnswer":{"@type":"Answer","text":"Yes, Fable and Mythos 5.2 are the same underlying model, but Fable includes additional safeguards around biology and cybersecurity topics to prevent misuse for creating dangerous weapons, while Mythos lacks those extra restrictions. Teams picking between Claude model variants for sensitive use cases can keep tabs on these distinctions via daily.dev."}}]} ```