In May, AI firm Anthropic introduced its Claude 4 family of models with a focus on improvements to coding, reasoning, and following instructions. Three months later, Anthropic is back with Claude Opus 4.1, which it says upgrades “agentic tasks, real-world coding, and reasoning.”
Anthropic says Claude Opus 4.1 improves software engineering accuracy to 74.5%. That compares to 62.3% with Claude Sonnet 3.7 and 72.5% with Claude Opus 4. More specifically, the updated model is better at “in-depth research and data analysis skills, especially around detail tracking and agentic search,” according to Anthropic.
Here’s a benchmark table provided by Anthropic’s blog post announcing Opus 4.1:
Anthropic’s latest model is out today for Claude customers and through Claude Code, through Anthropic’s API, Amazon Bedrock, and Google Cloud’s Vertex AI. For a deeper dive on Anthropic’s latest model, you can find the Claude Opus 4.1 system card here.
Meanwhile, Anthropic says on social media that it plans “to release substantially larger improvements to our models in the coming weeks,” so expect more model upgrades for Claude. OpenAI is also set to make news with announcements likely coming today and later this week.
Mac users can download Claude for macOS here. Anthropic also has iPhone and iPad apps for its Claude AI platform.