Anthropic Launches Claude Opus 4.7 with Xhigh Level
Get the Tech newsletter
Daily tech — startups, AI labs, chips, the launches that shape the next decade. Free.
- Anthropic released Claude Opus 4.7, calling it a notable improvement over Opus 4.6 in advanced software engineering.
- Claude Opus 4.7 adds a new “xhigh” effort level for higher‑capacity reasoning.
- Claude Opus 4.7 includes a new tokenizer and larger vision input capabilities.
- Claude Opus 4.7 shows big benchmark improvements across most tasks, but retains limited performance on long‑context and cyber benchmarks.
- Claude Opus 4.7 achieves graduate‑level reasoning comparable to GPT‑5.4 and Gemini 3.1 Pro, and is positioned as state‑of‑the‑art for agentic coding and computer use.
- Anthropic published a system‑card PDF for Claude Opus 4.7.
Why it matters: Developers and enterprises gain a more capable AI for software engineering, vision, and agentic coding, with benchmark gains across most tasks; however, the new “xhigh” tier likely raises token costs and the model’s deliberately limited long‑context and cyber abilities may restrict use in security‑sensitive or lengthy‑document scenarios.
Ask SkimNews
