Claude Opus 4.1

Anthropic·Aug 2025reasoningproprietary

Claude Opus 4.1 (August 2025) refined Claude Opus 4's agentic and coding abilities and was the top Claude until Sonnet 4.5 overtook it just weeks later. It is chiefly interesting today as a historical reference point: at $15/$75 it cost three times the current Opus pricing for materially less capability — a snapshot of how fast the 2025–2026 market moved.

Benchmark results

Where it shines

  • Stable legacy option for pinned enterprise deployments
  • Strong writing and refactoring quality for its era

Alternatives to Claude Opus 4.1

Frequently asked questions

How much does the Claude Opus 4.1 API cost?
Claude Opus 4.1 costs $15 per million input tokens and $75 per million output tokens, with cached input at $1.5 per million. A workload of 10M input and 1.5M output tokens per month costs about $262.50.
What is the context window of Claude Opus 4.1?
Claude Opus 4.1 supports a context window of 200,000 tokens (200K), with up to 32K output tokens per response.
Is Claude Opus 4.1 open source?
No — Claude Opus 4.1 is a proprietary model available through Anthropic's API and partner platforms.
What are the best alternatives to Claude Opus 4.1?
The closest alternatives by overall capability are Qwen3-Max, GPT-5.5, Gemini 3 Pro, Gemini 2.5 Pro. See the comparison pages for detailed head-to-head breakdowns.