Anthropic launches Claude Opus 5.5 at lower cost

Anthropic has launched Claude Opus 5.5, its latest flagship AI model for coding and knowledge work, marking the company's first major model release since Chief Executive Dario Amodei called for a slower pace in developing increasingly capable AI systems. The company prices Opus 5.5 at $4 per million input tokens and $20 per million output tokens, a 20 percent cut from Opus 5, which cost $5 and $25 respectively. Cache read costs have fallen 60 percent to $0.20 per million tokens, and output generation is more than 30 percent faster. Anthropic says total operating costs, accounting for both token prices and token usage, should be about 40 percent lower than Opus 5's on typical workloads. Anthropic says Opus 5.5 matches Claude Fable 5.1 performance on most tasks while running cheaper and faster. The company's published evaluations show the model scoring 66.4 on TerminalBench 4.0, 54.4 on FrontierCode, and 57.8 on CursorBench 4.0, alongside a 1,846 Elo on GDPvalAA v2.1, a benchmark measuring professional tasks across 44 occupations. Independent platform Artificial Analysis scored Opus 5.