Anthropic has announced the release of Claude Sonnet 4.5, which it describes as “the world’s best coding model” and a significant step forward in reasoning, mathematics, and agent development. The update positions Claude as one of the most advanced applied AI systems available, setting new benchmarks for developers, enterprises, and researchers.
A New Benchmark in AI Coding
“Claude Sonnet 4.5 is designed to be the strongest model for building complex agents and the most capable at using computers effectively,” Anthropic said. With measurable improvements in mathematical and reasoning tasks, the model aims to raise standards for applied AI across industries ranging from software development to enterprise automation.
Major Upgrades Across Claude Surfaces
The new version introduces enhancements across all Claude platforms:
Claude Code: A redesigned terminal interface, a VS Code extension, and a new checkpoints system allow developers to run large tasks with confidence and rewind to earlier code states instantly.
Claude App: Claude can now analyze data, generate files, and produce complete documents, presentations, and spreadsheets—all ready for download and further editing. These features are rolling out to all paid plans in preview.
Claude API: Developers gain access to tools for long-running, context-heavy agents. Features include context editing to clear outdated information automatically, and a memory tool that enables knowledge storage and recall beyond the model’s standard context window.
Research Preview: Imagine with Claude
Alongside the release, Anthropic has introduced an experimental feature called Imagine with Claude. Available exclusively to Claude Max users for five days, this research preview lets the model generate entire pieces of software on the fly—without prewritten code or predetermined functionality.
Availability and Pricing
Claude Sonnet 4.5 is now available globally through the Claude Developer Platform, Amazon Bedrock, and Google Cloud Vertex AI. Anthropic confirmed that pricing remains unchanged from Sonnet 4, making the upgraded model accessible at the same cost.
With Sonnet 4.5, Anthropic is signaling a push to consolidate its position at the forefront of applied AI, giving developers and enterprises more powerful tools to build agents, manage workflows, and unlock new productivity gains.
