Claude Opus 4.7 Released: Sharper Vision and Better Coding
3 min readAnthropic released Claude Opus 4.7 on April 16, 2026, delivering its most capable commercial model yet with a major leap in visual understanding and a new level of precision for complex coding tasks. At the same price as its predecessor, the update signals Anthropic’s confidence that raw performance gains can outpace cost concerns.
What Is Claude Opus 4.7?
Claude Opus 4.7 is Anthropic’s flagship reasoning model in the Claude 4 family, designed for demanding, multi-step tasks in software engineering, finance, and legal work. It builds on Opus 4.6 with two headline improvements: substantially sharper vision and more consistent performance on the hardest coding challenges developers throw at it.
Anthropic has positioned the Opus tier as the model for tasks where accuracy and reliability matter more than speed. Opus 4.7 continues that philosophy, adding finer control over how much effort the model applies to a given problem via a new xhigh effort level.
What Changed in This Release
The biggest visible upgrade is image resolution support. Opus 4.7 can now process images up to 2,576 pixels on the long edge, more than three times the maximum resolution of earlier Claude models. That improvement makes a real difference when analyzing technical diagrams, chemical structures, or any visual content where fine detail matters.
On the coding side, Anthropic reports a 13 percent lift on coding benchmarks and says the model resolves three times as many production-level tasks compared to Opus 4.6. The model also introduces better file-system-based memory, giving it more continuity across multi-session workflows rather than starting fresh each time.
Opus 4.7 achieved state-of-the-art results on GDPval-AA, a benchmark designed to measure economically valuable knowledge work in finance and legal domains. That positions it as a serious tool for enterprise teams, not just developers. According to Anthropic’s announcement, the model also shows improvements in honesty and resistance to prompt injection attacks.
Why the Same Price Matters
Pricing stays flat at $5 per million input tokens and $25 per million output tokens, matching Opus 4.6. In a market where competitors have either raised prices on new releases or tiered capabilities behind higher-cost plans, keeping pricing stable while improving performance is a notable competitive move.
This comes as Anthropic’s annualized revenue has surpassed $30 billion, up from $9 billion at the end of 2025. The company is betting that enterprise adoption at scale, driven by models like Opus 4.7, will sustain that trajectory. For API developers and enterprise customers already integrated with Claude, the upgrade path here is straightforward: same cost, meaningfully better results on the hardest tasks.
What to Watch Next
Anthropic has been previewing Claude Mythos to select partners, a model that reportedly sits above the Opus tier in capability. Opus 4.7 appears to be a polished stepping stone that keeps the commercial tier competitive while that next frontier model moves toward broader availability. If the pattern holds, Mythos could arrive publicly within the next few months.
For now, Opus 4.7 is generally available through the Claude API and on Claude.ai. Developers working on agentic pipelines, long-horizon coding tasks, or vision-heavy applications have the most to gain from upgrading.