Anthropic launched Claude Opus 5 on Thursday, releasing an AI model designed to deliver performance near its top-tier Fable 5 at half the price. Priced at $5 per million input tokens and $25 per million output tokens, the release expands enterprise options amid growing market cost scrutiny.
Anthropic has expanded its model lineup with the release of Claude Opus 5, positioning the new software as an everyday option for enterprises, knowledge workers, and developers. Arriving just two months after Opus 4.8 and following a string of summer model introductions, the release underscores a broader industry shift away from blockbuster rollouts and toward rapid iterations on cost, speed, and capability.
The company said Opus 5 approaches the capabilities of its top-tier Fable 5 model across many tasks while operating at half the cost. The model is priced at $5 per million input tokens and $25 per million output tokens, matching the price point of its predecessor according to Axios.
Pricing, Tiers, and Enterprise Integration
Opus 5 is immediately rolling out across multiple subscription levels and developer platforms. It serves as the default model on Claude Max, and functions as the strongest available model on the $20-per-month Claude Pro tier.
Beyond Anthropic’s native apps, GitHub announced that Claude Opus 5 is now available in GitHub Copilot for Copilot Pro+, Max, Business, and Enterprise users. Enterprise and Business plan administrators must manually enable the Opus 5 policy within their Copilot settings to grant access across tools like Visual Studio Code, Xcode, and JetBrains.
Performance Benchmarks and Scientific Research
In developer evaluations, Opus 5 establishes new performance marks for agentic coding and knowledge work. On the OSWorld 2.0 computer use benchmark, the model outperforms every other model at any given cost, surpassing Fable 5’s best result at just over a third of the price according to The New Stack. On ARC-AGI 3, which tests problem-solving on novel tasks, the model scored three times as high as the next-best competitor.
Anthropic also highlighted gains in scientific research evaluations. The company called Opus 5 its most capable generally available model for scientific research, noting substantial improvements in organic chemistry tasks—such as inferring molecular structures from spectroscopy data—and protein-related evaluations detailed in company release notes.
Safety Adjustments and Government Testing
The release arrives against a backdrop of heightened regulatory scrutiny and industry security incidents. Following export control directives and negotiations earlier in the summer regarding advanced model capabilities, Anthropic emphasized the safety posture of its latest software.
The company described Opus 5 as the most aligned Opus model and the least susceptible to being tricked into misuse. While the Mythos 5 model remains the stronger model for offensive cybersecurity and biological research, Opus 5 retains a similar portfolio of biological safeguards to Opus 4.8.
Anthropic confirmed that government stakeholders evaluated the software prior to release noted by CNBC.
“We continue to work with our government partners to conduct their own independent testing of our models. This includes Opus 5,”
Danielle Ghiglieri
