Anthropic Launches Claude Opus 5.5, Cuts Costs 40% and Scraps 5-Hour Usage Caps
Anthropic launched Claude Opus 5.5 on Tuesday, introducing a new model line as the AI company looks to compete with rivals, including OpenAI, on performance, speed, and cost. Anthropic said Opus 5.5 delivers performance approaching its previous top-tier Claude models while requiring less compute. The company said the model costs 40% less on typical workloads at default settings than Claude Opus 5 and generates responses faster. The launch also comes with changes to Claude’s subscription plans. Anthropic said it is removing five-hour usage caps for Pro, Max, Team and seat-based Enterprise subscribers and adding a rate-limit reset feature. The company highlighted safety improvements alongside the performance gains. Anthropic said external groups including Frontier Design and METR evaluated Opus 5.5 before launch, while an automated behavioral audit produced the strongest results the company has recorded. Anthropic also said the model is less likely to take difficult-to-reverse actions than newer models and is more resistant to prompt injection than Opus 5. Opus 5.5’s capabilities in areas such as biology and cybersecurity will include safeguards similar to those used for Claude Fable
5 on Tuesday, introducing a new model line as the AI company looks to compete with rivals, including OpenAI, on performance, speed, and cost. 5 delivers performance approaching its previous top-tier Claude models while requiring less compute. The company said the model costs 40% less on typical workloads at default settings than Claude Opus 5 and generates responses faster. The launch also comes with changes to Claude’s subscription plans.
Anthropic said it is removing five-hour usage caps for Pro, Max, Team and seat-based Enterprise subscribers and adding a rate-limit reset feature. The company highlighted safety improvements alongside the performance gains. 5 before launch, while an automated behavioral audit produced the strongest results the company has recorded. Anthropic also said the model is less likely to take difficult-to-reverse actions than newer models and is more resistant to prompt injection than Opus 5.
5’s capabilities in areas such as biology and cybersecurity will include safeguards similar to those used for Claude Fable 5. Vetted organizations can apply to the company’s Life Sciences Verification Program, while Anthropic plans to expand access to its Cyber Verification Program in the coming weeks. 5 performing at or near the top in categories including agentic coding, computer use and knowledge work. Anthropic cautioned that relatively small differences in benchmark scores do not necessarily translate into meaningful differences in real-world performance.
20 per million tokens. A faster mode is also available at higher token rates. Early testers cited by Anthropic reported completing large-scale code migrations and audits in hours rather than days. 5 can identify software performance bottlenecks while making fewer changes to application behavior than Opus 5.
5 in the coming weeks, extending the new model family beyond its flagship Opus offering. Photo: Shutterstock Read Also: OpenAI Loses Key Design Leader After Shaping ChatGPT And Codex