Anthropic announced the release of Claude Opus 4.1, a refined successor to Claude Opus 4 that delivers notable improvements in agentic workflows, real‑world coding, reasoning quality, and instruction‑following.
Claude Opus 4.1 is now available to paid Claude users, Claude Code subscribers, and developers via the Anthropic API, Amazon Bedrock, and Google Cloud’s Vertex AI, priced identically to Opus 4.
Key Features & Performance
- Anthropic positions Claude Opus 4.1 as a drop‑in replacement for Opus 4, offering superior precision in real‑world coding and agentic tasks.
- The model achieves 5% on SWE‑bench Verified, improving over the 72.5% from Opus 4 and establishing new industry leadership in software engineering evaluation.
- Significant gains observed in multi‑file code refactoring and debugging within large codebases, particularly noted by GitHub, Rakuten, and Windsurf users.
- Enhanced performance in agentic search, data analysis, and research capabilities, enabling long‑duration, detail‑oriented workflows.
Also Read: Google DeepMind Announces General Availability of Gemini2.5 Deep Think for AIUltra Subscribers
Customer Feedback
“Within a large codebase, Claude Opus 4.1 faithfully adhered to instructions and pinpointed the exact spot requiring correction–without making unnecessary adjustments or introducing new bugs. Our team prefers this attention to detail for everyday debugging tasks. We also saw impressive efficiency gains: up to 50% faster task completion with 45% fewer tool uses.” – Kenta Naruse, ML Engineer at Rakuten
“In our benchmark for junior developers, Claude Opus 4.1 shows roughly the same performance leap as the jump from Sonnet 3.7 to Sonnet 4. Opus 4.1 delivers a one standard deviation improvement over Opus 4. Very excited for our users to feel this difference.”– Jeff Wang, CEO of Windsurf
“Early testing with Claude Opus 4.1 demonstrates improvements across most capabilities relative to Opus 4, with particularly notable performance gains in multi‑file code refactoring.”
Use Cases & Availability
Claude Opus 4.1 is Anthropic’s most capable model to date, powering complex AI agent applications, extensive coding tasks (including multi‑step, thousands‑of‑step workflows), agentic search across large data sources, and refined content generation with human‑quality prose. It also supports hybrid reasoning via instant responses or extended step‑by‑step thinking, with summaries instead of full reasoning chains for usability and interpretability.
The release is available across multiple deployment channels:
- Claude Pro, Max, Team, and Enterprise tiers
- Claude Code terminal interface
- Anthropic API
- Amazon Bedrock (US West, US East regions)
- Google Cloud Vertex AI
