Anthropic unveils Claude Opus 4.1 — a steadier hand for hard work

Anthropic introduced Claude Opus 4.1, an update to its flagship model. The promise is simple: better results on long, real-world tasks. It plans, checks its own steps, and follows through. Code reads clearer. Fixes land faster. Research flows with fewer stumbles.

Opus 4.1 is a drop-in replacement for Opus 4. No new knobs to turn. The price stays the same. Paid Claude users can try it at once. It is also live on the API. You can reach it through Amazon Bedrock and Google Cloud’s Vertex AI. Developers will meet it inside Claude Code. GitHub Copilot has opened a public preview as well.

Anthropic frames this release as a step, not a finish line. Larger gains are on the way, the company says. For teams that build and ship daily, the change matters now. Fewer rewrites. Fewer brittle scripts. More steady work from prompt to pull request. In short, a model that feels less like a demo and more like a colleague you can trust.