{"version":1,"type":"story","url":"https://digestai.news/story/claude-opus-5-5-tops-benchmark-over-openais-astra-and-fable-5-1","json":"https://digestai.news/story/claude-opus-5-5-tops-benchmark-over-openais-astra-and-fable-5-1.json","markdown":"https://digestai.news/story/claude-opus-5-5-tops-benchmark-over-openais-astra-and-fable-5-1.md","slug":"claude-opus-5-5-tops-benchmark-over-openais-astra-and-fable-5-1","headline":"Claude Opus 5.5 tops benchmark over OpenAI’s Astra and Fable 5.1","summary":"Anthropic’s new model, **Claude Opus 5.5**, ranked first in the **Intelligence Index** benchmark by *Artificial Analysis*, surpassing OpenAI’s **GPT-6 Astra** and Anthropic’s own **Fable 5.1**. The index aggregates scores from ten tests, including programming and knowledge-based queries. Opus scored **58 points**, while Astra and Fable both scored **53**. However, the analysis notes that Astra uses fewer tokens at a lower cost, making it more economical for cost-sensitive users despite slightly lower performance in some areas like scientific programming, where Opus leads by **11 percentage points** over Astra and **4 points** over Fable.\n\nThe benchmark tested models at their highest performance settings. While Opus excels in performance, the analysis advises users to weigh cost efficiency—especially since Astra’s token pricing is cheaper than Opus’s, though both remain expensive. Fable 5.1, an older Anthropic model, underperformed in both metrics.","keyPoints":["Claude Opus 5.5 scored **58** in the Intelligence Index, ahead of OpenAI’s Astra (**53**) and Fable 5.1 (**53**)","Opus leads Astra by **11 percentage points** in scientific programming but costs more per token","Artificial Analysis recommends Astra for cost-conscious users despite lower performance in some tasks"],"whyItMatters":"The benchmark shift highlights Opus 5.5’s performance edge but underscores cost as a key differentiator in AI model selection, influencing enterprise and developer decisions.","category":{"slug":"models","name":"Generative AI & Models","url":"https://digestai.news/category/models"},"entities":{"companies":["Anthropic","OpenAI","Artificial Analysis"],"models":["Claude Opus 5.5","GPT-6 Astra","Fable 5.1"],"people":[]},"firstPublishedAt":"2026-09-25T04:42:00Z","updatedAt":"2026-09-25T04:42:00Z","sourceCount":1,"hasPrimarySource":false,"sources":[{"outlet":"heise.de","title":"External benchmarks: Claude Opus 5.5 overtakes OpenAI’s Astra and Fable 5.1","url":"https://heise.de/en/news/External-benchmarks-Claude-Opus-5-5-overtakes-OpenAI-s-Astra-and-Fable-5-1-11466259.html","publishedAt":"2026-09-25T04:42:00Z","type":"press","primary":false,"lead":true}],"sourceNotes":null,"discussions":[],"thread":{"title":"Anthropic's Cost Effective AI Model Dominates Benchmarks","url":"https://digestai.news/thread/claude-opus-5-5-costs-60-less-than-fable-5-1-for-writing-tasks","storyCount":2},"cite":{"text":"Digest AI, \"Claude Opus 5.5 tops benchmark over OpenAI’s Astra and Fable 5.1\", 25 September 2026, https://digestai.news/story/claude-opus-5-5-tops-benchmark-over-openais-astra-and-fable-5-1","publisher":"Digest AI","title":"Claude Opus 5.5 tops benchmark over OpenAI’s Astra and Fable 5.1","datePublished":"2026-09-25T04:42:00Z","url":"https://digestai.news/story/claude-opus-5-5-tops-benchmark-over-openais-astra-and-fable-5-1"},"generatedBy":"Written by Digest AI's editorial model from the linked sources; the sources are the record.","license":"Headlines, digests and key points are written by Digest AI and may be quoted with a link to the story page. Linked articles belong to their publishers. Terms: https://digestai.news/terms#reuse"}