Claude Opus 4.7

Anthropic's latest flagship. AA Intelligence Index: 57 (max reasoning effort), tied with Gemini 3.1 Pro and GPT-5.4. First Claude model with high-resolution image support (2576px / 3.75MP). New tokenizer for improved performance. Handles complex, long-running tasks with rigor and consistency, pays precise attention to instructions, and devises ways to verify its own outputs.

Notable improvement over Opus 4.6 in advanced software engineering, especially on the hardest tasks. Released alongside Claude Mythos (withheld from public release for cybersecurity safety). Same pricing as Opus 4.6: $5/M input, $25/M output. Available across Claude products, API, Amazon Bedrock, Google Cloud Vertex AI, and Microsoft Foundry.