Introducing Claude Opus 4.7ProductApr 16, 2026Our latest Opus model brings stronger performance across coding, agents, vision, and multi-step tasks, with greater thoroughness and consistency on the work that matters most.

Anthropic announced the general availability of Claude Opus 4.7 on April 16, 2026, its latest model aimed at professional and enterprise users. The update brings significant performance improvements over its predecessor, Opus 4.6, particularly in advanced software engineering, vision capabilities, and the execution of complex, multi-step tasks. According to the company, the model is designed to handle difficult coding assignments with greater autonomy and consistency, making it a more reliable tool for developers. The model is immediately available via the Claude API and on cloud platforms including Amazon Bedrock, Google Cloud’s Vertex AI, and Microsoft Foundry.

Technical Details and Safeguards

While Anthropic positions Opus 4.7 as more capable than Opus 4.6, it is noted to be less powerful than the limited-release Claude Mythos Preview. The new model serves as a testbed for new cybersecurity safeguards, with its cyber capabilities having been differentially reduced during training. These safeguards are designed to automatically block prohibited or high-risk requests. Pricing for Opus 4.7 remains unchanged from its predecessor at $5 per million input tokens and $25 per million output tokens.

• Enhanced Coding: Multiple early-access partners reported double-digit performance lifts on coding benchmarks like CursorBench and Rakuten-SWE-Bench.
• Improved Vision: The model processes images at a higher resolution, aiding in tasks like interpreting complex technical diagrams and chemical structures.
• Agentic Consistency: Testers highlighted better reliability in long-running autonomous tasks, with improved error recovery and resistance to looping.
• Cybersecurity Controls: The model includes new safeguards and its advanced cyber capabilities are restricted. Legitimate security professionals can apply for access via a new Cyber Verification Program.

Market Impact and Enterprise Adoption

The release of Claude Opus 4.7 signals a focus on making AI agents more practical for production enterprise workflows. Feedback from partners such as Hex, Replit, and Harvey emphasizes the model's increased reliability and deeper reasoning. Instead of just raw performance, users noted its ability to catch its own logical faults, push back on user suggestions with its own opinionated perspective, and handle ambiguous tasks with greater accuracy. This emphasis on consistency and dependable execution for long-horizon work addresses a key barrier to broader enterprise adoption of autonomous AI systems, positioning the model less as a raw tool and more as a capable digital coworker for technical teams.

Anthropic's strategy with Claude Opus 4.7 indicates a market shift from prioritizing peak benchmark performance to delivering the operational consistency and built-in safeguards required for deploying autonomous agents in real-world, high-stakes enterprise environments.