Anthropic announced its new AI model, "Claude Opus 5.5," on September 22. It reduces execution costs by 40% compared to the previous Opus 5 and boosts output speed by over 30%. API pricing has also been reduced by 20% from Opus 5, now costing $4 per million input tokens and $20 per million output tokens.
Its capabilities in coding and complex tasks have been enhanced. According to stated figures, it has surpassed previous models in multiple evaluations, including scoring 66.4% on "Terminal-Bench 4.0."
OpenAI's latest models were also presented as comparison targets. Benchmark comparisons with GPT-6 Astra and GPT-5.6 Sol were published, showing Opus 5.5 at 66.4% and GPT-6 Astra at 57.9% on Terminal-Bench 4.0. Conversely, on Terminal-Bench-Science, designed for scientific research, Astra scored 64.6% while Opus 5.5 achieved 58.7%.
Similar to Anthropic, OpenAI also announced "GPT-6 Sol" and "GPT-6 Luna" on the same day, September 22. These models deploy the technology introduced in the flagship GPT-6 Astra into faster and lower-cost versions.
OpenAI also highlighted price reductions, with Sol priced at $2 per million input tokens and $10 per million output tokens, and Luna at $0.10 per million input tokens and $0.50 per million output tokens.
Anthropic also plans to release "Claude Sonnet 5.5" and "Claude Haiku 5.5" in the coming weeks.
