Anthropic drops Claude 4.6 Opus — a new state-of-the-art in reasoning and coding. Here's what changed, what it costs, and whether to upgrade.
Anthropic has released Claude 4.6 Opus, the newest and most capable model in the Claude family. In both internal benchmarks and third-party evaluations, Opus 4.6 outperforms GPT-4o and Gemini 1.5 Pro on coding, reasoning, and long-context tasks.
The biggest upgrade in Opus 4.6 is a 40% improvement on SWE-Bench Verified (real-world software engineering tasks), pushing its score to 72.5% — a new state of the art. On GPQA Diamond (graduate-level reasoning), it scores 71.9%.
Context window remains at 200K tokens. Pricing is $15 per million input tokens and $75 per million output tokens — unchanged from Opus 4.
Claude Sonnet 4.6 also received an update, offering much of the Opus reasoning improvement at a lower price point ($3/M input, $15/M output). For most production API applications, Sonnet 4.6 is the right choice.
Existing API keys and integrations require no changes — update the model parameter to claude-opus-4-6 or claude-sonnet-4-6 and you are done.
Weekly AI tool reviews, news digests, and how-to guides.
Join 12,000+ builders