Anthropic Claude 4: Coding & Reasoning

Anthropic has launched Claude Opus 4 and Claude Sonnet 4, the latest in hybrid-reasoning AI models. These models are designed for coding and solving difficult problems.

Claude Opus 4

Anthropic says Claude Opus 4 is their most powerful AI model and can work on long tasks for several hours. In customer testing, Opus 4 worked autonomously for seven hours. The company calls it the “best coding model in the world,” and says it outperformed Google’s Gemini 2.5 Pro, OpenAI’s o3 reasoning, and GPT-4.1 in coding tasks and using web search.

Claude Sonnet 4

Claude Sonnet 4 is more affordable and efficient and is intended for general tasks. It supersedes the 3.7 Sonnet model from February. Anthropic states that Sonnet 4 delivers better coding and reasoning with more precise responses. The models are also 65 percent less likely to take shortcuts compared to 3.7 Sonnet. They also store key information better for long-term tasks when given local file access.

New Features

Both Claude 4 models have “thinking summaries,” which condense the chatbots’ reasoning. There is also an “extended thinking” beta feature that allows users to switch between modes for reasoning or using tools.

Availability

Claude Opus 4 and Sonnet 4 are available on the Anthropic API, Amazon Bedrock, and Google Cloud’s Vertex AI platform. Both models are included in paid Claude plans alongside the extended thinking beta feature. Free users can access Claude Sonnet 4 for now.

Anthropic’s Claude Code agentic command-line tool is now generally available after its preview in February. Anthropic is also planning “more frequent model updates” to keep up with competition from OpenAI, Google, and Meta.