Anthropic Launches Claude Opus 4.7 With Self Verification for Autonomous Agents

Apr 16, 2026 · Updated Apr 25, 2026

Anthropic launched Claude Opus 4.7, a model iteration specifically engineered for long-horizon agentic workflows (autonomous systems that complete multi-step tasks over hours). It introduces self-verification, where the model identifies logical faults during planning and validates its own results. The update also includes a 3x vision resolution increase for reading diagrams.

This release addresses the reliability gap in autonomous agents, which often struggle with "looping" or stalling during tasks. By improving instruction following and adding a new xhigh effort level (a setting that controls the reasoning-to-latency tradeoff), the model enables more consistent performance on asynchronous work like CI/CD pipelines.

The model is available via the Claude API, Amazon Bedrock, and Google Cloud. In Claude Code, the new /ultrareview command uses this model to flag bugs and design issues that a careful reviewer would catch. This enables more autonomous, multi-step coding tasks with fewer interruptions.

View the full update on anthropic.com

Claude

@claudeaiApr 16

Introducing Claude Opus 4.7, our most capable Opus model yet. It handles long-running tasks with more rigor, follows instructions more precisely, and verifies its own outputs before reporting back. You can hand off your hardest work with less supervision. https://t.co/PtlRdpQcG5

8.3k66k

View on X