Anthropic Releases Claude Opus 5, Matching Fable 5 Performance at Half the Price
AI company Anthropic has launched its latest flagship model, Claude Opus 5. The model demonstrates performance comparable to Fable 5 but is priced at half the cost, setting new benchmarks in various evaluations.
Anthropic has released Claude Opus 5, the latest iteration of its flagship AI model. The company reports that Opus 5 achieves performance levels close to its higher-tier Fable 5 model, while offering a significantly lower price point.
Positioned for demanding daily use, Opus 5 is now the default model for Claude Max and the most capable option for Claude Pro subscribers. Anthropic states the model has set new records in several benchmarks, including Frontier-Bench and ARC-AGI 3, excelling in coding, knowledge work, and scientific research tasks.
In coding assessments like Frontier-Bench v0.1, Opus 5 reportedly surpassed other models and more than doubled task performance compared to Opus 4.8, with reduced costs per task. Similarly, in CursorBench 3.2, it closely matched Fable 5's peak performance at half the cost.
The model also shows advancements in scientific applications, performing notably better in areas like structural biology and organic chemistry tasks compared to its predecessor. Anthropic highlighted its improved ability to self-correct and iterate on solutions, demonstrating complex problem-solving capabilities in real-world scenarios.
Opus 5 is priced identically to Opus 4.8: $5 per million input tokens and $25 per million output tokens. Anthropic has also introduced an accelerated "Opus 5 Fast" mode, doubling speed for double the price, and new features for developers aimed at improving workflow and error handling.