Back to News
AI
Anthropic claims its new Claude Opus 5 delivers near-Fable 5 performance at half the token price
The Decoder·July 24, 2026·1 min read
AI Summary
Anthropic's Claude Opus 5 model achieves performance comparable to Fable 5 while costing half as much in tokens. It scored 30.2 percent on the ARC-AGI-3 benchmark, significantly outperforming GPT-5.6 Sol by nearly four times.
From the source
Anthropic's new flagship model Claude Opus 5 posts top scores in coding and knowledge work at half of Fable 5's token rates. On ARC-AGI-3, a benchmark for novel problem-solving, Opus 5 hits 30.2 percent, nearly four times higher than GPT-5.6 Sol. The article Anthropic claims its new Claude Opus 5 delivers near-Fable 5 performance at half the token price appeared first on The Decoder.
The full text couldn't be loaded here (the source may require a subscription).
View original at The DecoderKeep reading
Evaluating Book Reports: Emphasis on Independent Thought, Judgment, and Expression, Incorporating Oral Presentations and Discussions, Amid Easy Use of Generative AItopics.smt.docomo.ne.jp · 3h agoWarning shot or publicity stunt - how worried should we be about the OpenAI hack? - BBCbbc.co.uk · 3h agoPresident Lee Jae-myung Meets Anthropic CEO, Calls Korea 'Asia's Optimal AI Partner'kyeonggi.com · 3h agoKim Seong-joo's AI Photo Recipe: Wow! Sharp Evaluation by Artificial Intelligencehani.co.kr · 3h ago
Was this useful?