Claude Opus 4.7 on AI Gateway
By Steven Van ·
Anthropic's Opus 4.7 adds task budgets, an 'xhigh' effort level, and better memory and chart-reading for agents on Vercel's AI Gateway.
Anthropic's Claude Opus 4.7 is now available on Vercel's AI Gateway. The model is built for long-running, asynchronous agents and handles complex, multi-step tasks with reliable agentic execution, with gains on knowledge-worker tasks, particularly where it needs to visually verify its own outputs.
- Stronger programmatic tool-calling with image-processing libraries for analyzing charts and figures, including pixel-level data transcription, plus high-resolution image support for computer use, screenshot understanding, and document analysis.
- Improved memory: agents that keep a structured memory store across turns get more reliable recall and fewer dropped facts without extra prompting.
- A new 'xhigh' effort level, used by setting the model to anthropic/claude-opus-4.7 in the AI SDK.
- A task budgets feature: setting a total token budget for an agentic turn via taskBudget shows the model a countdown of remaining tokens, which it uses to prioritize work, plan ahead, and wind down gracefully as the budget is consumed.
- Thinking content is now omitted by default for Opus 4.7; set display to 'summarized' to receive it.
Vercel
The platform for frontend developers — deploy, preview, and scale web apps and AI agents with zero config.
View Vercel →