Anthropic has launched Claude Opus 4.7, the latest upgrade to its flagship model family, now generally available across all Claude products and its API. The release marks a meaningful step up from Opus 4.6, with the biggest gains in advanced software engineering, vision quality, and long-running autonomous task performance.

Claude Opus 4.7 is powered by an updated tokenizer and a new xhigh effort level, giving developers finer control over the tradeoff between reasoning depth and response speed. Users who tried the model in early access report being able to hand off complex, multi-step coding work that previously needed close supervision. The model catches logical faults during planning, verifies its own outputs before reporting back, and maintains consistency across long agentic runs.

What Changed in Claude Opus 4.7:

• Advanced software engineering: Coding performance improved across the board, with early testers reporting double-digit gains on internal benchmarks. On CursorBench, Opus 4.7 cleared 70% of tasks compared to Opus 4.6's 58%.
• Stronger vision: The model now accepts images up to 2,576 pixels on the long edge, more than three times the resolution of prior Claude models, enabling finer detail in screenshots, diagrams, and multimodal workflows.
• Better instruction following: Opus 4.7 interprets prompts more literally and thoroughly. Prompts written for earlier models may need re-tuning, as the model now executes instructions exactly as written.
• Improved memory: The model uses file system-based memory more effectively, carrying context across long, multi-session work without needing full re-briefing on each task.
• Real-world task performance: It is now state-of-the-art on the GDPval-AA benchmark, a third-party evaluation covering finance, legal, and knowledge work.

New Features Shipping Alongside Opus 4.7:

• xhigh effort level: A new setting between high and max for finer reasoning control. Claude Code now defaults to xhigh for all plans.
• /ultrareview in Claude Code: A dedicated code review command that flags bugs and design issues. Pro and Max users get three free ultrareviews.
• Task budgets (public beta): Developers can guide Claude's token spend across longer agentic runs via the API.
• Auto mode for Max users: Claude makes permissions decisions autonomously, enabling longer tasks with fewer interruptions.

Cybersecurity Safeguards:

Opus 4.7 is the first Claude model to ship with real-time cybersecurity safeguards that automatically detect and block prohibited or high-risk uses. Security professionals can apply to Anthropic's Cyber Verification Program for legitimate access.

Pricing stays the same as Opus 4.6: $5 per million input tokens and $25 per million output tokens. Developers can access it via claude-opus-4-7 on the Claude API, Amazon Bedrock, Google Cloud's Vertex AI, and Microsoft Foundry.

Also Read

• Anthropic Launches Claude Design for Teams to Create Prototypes, Decks, and Landing Pages
• Anthropic Launched Claude Opus 4.5 — New Flagship Model for Coding and Complex AI Workflows
• Microsoft Copilot Adds Multi-Model AI Comparison
• Google Launches Gemma 4, Its Most Capable Open Model Family
• OpenAI Closes Record $122 Billion Funding Round at $852 Billion Valuation