We've launched Claude Opus 5 (claude-opus-5), a step-change improvement over Claude Opus 4.8
By Steven Van ·
We've launched Claude Opus 5 (claude-opus-5), a step-change improvement over Claude Opus 4.8.
Anthropic has launched Claude Opus 5 (claude-opus-5) on the Anthropic platform, calling it a step-change improvement over Claude Opus 4.8. Pricing holds at $5/$25 per MTok, the same as Opus 4.8, and the model is available on the Claude API, Claude in Amazon Bedrock, Claude Platform on AWS, Claude on Google Cloud, and Claude in Microsoft Foundry.
- Context window is 1M tokens by default and at maximum, with 128k max output tokens.
- Thinking is on by default. Disabling it is only allowed at effort high or below: setting thinking to disabled with effort xhigh or max now returns a 400 error, a breaking change from Opus 4.8.
- Effort is the primary control for steering the model, with the full ladder from low to max, and max meant for capability-critical work.
- Mid-conversation tool changes are in beta on Opus 5, alongside Claude Fable 5, Claude Mythos 5, and Opus 4.8: tools can be added or removed between turns of a conversation while preserving the prompt cache.
Anthropic
Anthropic is an AI safety and research company that's working to build reliable, interpretable, and steerable AI systems.
View Anthropic →