Anthropic's Claude Opus 5 delivers near-Fable 5 performance at half the token price

Anthropic launched Claude Opus 5 as its new flagship model, targeting near-Fable 5 performance at much lower token prices.

The model keeps a 1 million-token context window and is now the default for Claude Max, with Claude Pro users able to choose it as the top option.

Claude Opus 5 shows strong results in coding and knowledge-work benchmarks, though performance varies by effort setting and it does not top every coding test.

The release tightens the cost-performance gap in frontier AI and increases pressure on premium and rival models, especially for coding and agentic workloads.