Anthropic's Claude Opus 4.1 Boosts Coding, Outshines Rivals

• Claude Opus 4.1 achieves 74.5% on SWE-bench Verified, surpassing Opus 4’s 72.5% score.
• Outperforms OpenAI o3, Gemini 2.5 Pro in Agentic Coding, Multilingual Q&A benchmarks.
• Available to Claude Pro ($20/month), Max ($100/month), via API, Bedrock, and Vertex AI.

Anthropic has launched Claude Opus 4.1, the successor to Claude Opus 4, boasting significant improvements in coding, reasoning, and agentic tasks. According to Anthropic’s blog, Opus 4.1 “improves Claude’s in-depth research and data analysis skills, especially around detail tracking and agentic search.”

The model scores impressively at 74.5% on the SWE-bench Verified, a benchmark that measures AI performance on realistic programs of real-world software engineering using GitHub repositories, higher than the 72.5% that the Opus 4 achieved.

Outperforming competitors like OpenAI’s o3 and Gemini 2.5 Pro in benchmarks such as Agentic Coding and Multilingual Q&A, Opus 4.1 solidifies Anthropic’s edge in coding-focused AI. However, it trails rivals in tasks like visual reasoning and high school math.

Also Read: DeepMind's Genie 3 Creates Real-time 3D Interactive Worlds

The model is available to Claude Pro subscribers ($20/month), Claude Max users ($100/month), and through Claude Code, as well as via API, Amazon Bedrock, and Google Cloud’s Vertex AI, maintaining the same pricing as Opus 4.

Launched shortly after Claude Opus 4’s debut in late May, Opus 4.1 comes amid competitive tensions. Last week, Anthropic revoked OpenAI’s access to Claude Code after discovering its use ahead of OpenAI’s anticipated GPT-5 launch, rumored to enhance coding capabilities to rival Claude’s popularity among developers.

This new release reflects Anthropic’s need to drive AI for large, complicated tasks, such as software engineering, and exists within a crowded market with major industry competitors.