Anthropic has officially launched Claude Opus 4.6, its most advanced large language model, on Thursday, February 5, 2026. The new model introduces substantial enhancements across various benchmarks and is particularly geared towards complex financial research, enterprise workflows, and sophisticated software development.
According to Anthropic, Claude Opus 4.6 is designed to scrutinize company data, regulatory filings, and market information to generate detailed financial analyses that would typically take days for a human to complete. An Anthropic spokesperson highlighted its ability to achieve “production-ready quality on the first try” for documents, spreadsheets, and presentations, thereby reducing iterative work.
A key feature of Opus 4.6 is its expanded one-million-token context window and an increased output capacity of up to 128,000 tokens. The model also supports agent teams within Claude code, enabling tasks to be executed in parallel. Pricing for Claude Opus 4.6 remains consistent with its predecessor, at $5 per million input tokens and $25 per million output tokens.
In terms of performance, Anthropic states that Opus 4.6 marks a “step change” for enterprise users. In internal evaluations focused on real-world finance use cases, Claude Opus 4.6 demonstrated a significant improvement of over 23 percentage points compared to Claude Sonnet 4.5. While the model shows considerable gains in general coding benchmarks and agentic tool use, some minor regressions were noted in SWE-bench verified tests and the MCP Atlas benchmark.
The model’s release also includes updates to existing products, such as an improved “Claude in Excel” for handling longer and more complex multi-step tasks, and the introduction of “Claude in PowerPoint” as a research preview beta for native presentation building. Anthropic classifies Opus 4.6 as a “Level 3” model on its four-point safety scale, indicating that its power introduces “significantly higher risk.”
Claude Opus 4.6 is available to developers and enterprises through the Anthropic API, AWS Bedrock, and Google Vertex AI.