Anthropic's Claude Opus 5 Model Tops AI Leaderboard and Cuts Costs

Anthropic's latest large‑language model, Claude Opus 5, achieved the number‑two position on the Agent Arena leaderboard, just behind the company's own Claude Fable 5. The Opus 5 Max variant recorded an 11.88% net improvement score and a 17.56% confirmed success rate, while the Opus 5 High version posted similar metrics, securing the third spot.

The company markets Opus 5 as delivering performance close to Fable 5 at roughly half the price. Both Opus 5 Max and High are now the default models on Claude Max plans and are available on Claude Pro. Pricing remains unchanged at $5 per million input tokens and $25 per million output tokens, with additional discounts for caching and batch processing. Anthropic emphasizes the model’s cost‑adjusted competitiveness for coding, knowledge‑work, and long‑horizon tasks, positioning it as a more affordable tier for developers who need strong AI capabilities without the premium cost of Fable 5.