Anthropic launches Claude Opus 4.8 with improved reasoning capabilities
Early testing shows the new model is more accurate and less likely to make unsupported claims, improving trust in agentic AI systems.
Anthropic has introduced Claude Opus 4.8, an upgraded version of its flagship AI model, with improvements across coding, agentic tasks, reasoning, and practical knowledge work.
The company said the model builds on Claude Opus 4.7 and is available at the same regular pricing. Anthropic also said that fast mode for Opus 4.8 can run 2.5 times as fast and is now 3 times cheaper than fast mode for previous models.
A key focus of the release is reliability. Anthropic said early testers found Opus 4.8 sharper in judgement when performing agentic tasks, more likely to flag uncertainty, and less likely to make unsupported claims. The company’s evaluations also found the model to be around four times less likely than its predecessor to leave flaws in its own code unremarked.
New features include dynamic workflows in Claude Code, available in research preview, allowing Claude to plan and run hundreds of parallel subagents in a single session for large-scale tasks. Anthropic said the feature can support codebase-scale migrations across hundreds of thousands of lines of code.
Users on claude.ai and Claude Cowork can also control how much effort Claude applies to a response. Higher effort settings are designed to improve quality for difficult tasks, while lower effort settings allow faster responses and slower use of rate limits.
Anthropic also reported stronger alignment results for Opus 4.8 compared with Opus 4.7. Its alignment assessment found lower rates of misaligned behaviour, such as deception or misuse of cooperation, and stronger support for user autonomy and user interests.
The model is available across Anthropic’s platforms, and developers can access it through the Claude API using the claude-opus-4-8 model name.
Why does it matter?
Claude Opus 4.8 shows how frontier AI competition is moving beyond benchmark performance towards reliability in professional workflows. Features such as effort control, dynamic workflows, cheaper fast mode, and stronger agentic task performance point to a market shift in which AI systems are expected to manage longer, more complex work in coding, research, analysis, and enterprise operations while giving users greater control over cost, speed, and reasoning depth.
Would you like to learn more about AI, tech, and digital diplomacy? If so, ask our chatbot!