Claude Sonnet 4 Receives Massive Update to Support 1M Tokens

Anthropic has supercharged Claude Sonnet 4. The hybrid AI model, now in beta, can handle up to 1 million tokens of context, a fivefold jump from previous versions. That’s enough to process entire codebases with 75,000+ lines or dozens of research papers in a single request.

Claude Sonnet 4 updates, availability,and pricing

Anthropic’s blog about the Claude Sonnet 4 boost says it is a “5x increase” that unlocks massive new possibilities for developers and researchers; the update is meant specifically for those with large-scale AI workloads. This news comes a week after the Anthropic team unveiled Claude Opus 4.1 and just days after rolling out new features to Claude Code.

Anthropic provided these availability details about Claude Sonnet 4: “Long context support for Sonnet 4 is now in public beta on the Anthropic API and in Amazon Bedrock, with Google Cloud’s Vertex AI coming soon.”

The team at Anthropic has also adjusted its model’s pricing for prompts that exceed 200K tokens.

• Prompts ≤ 200K:  $3 / MTok (Input) or $15 / MTok (Output)
• Prompts > 200K: $6 / MTok (Input) or $22.50 / MTok (Output)

Claude’s expanded pricing is due to the increased computational requirements of long context support; however, users can combine prompt caching or batch processing functionality to reduce their overall cost.

New Claude Sonnet 4 use cases

With its expanded 1 million–token context, Claude Sonnet 4 can now process projects that previously had to be split into multiple prompts. This “long context support” means the model can work through sprawling datasets, multi-part technical documents, and interconnected workflows without losing track of the bigger picture.

New use cases for Claude Sonnet 4 include:

• Code processing and analysis: Process and analyze entire codebases, complete with source files, technical documentation, and test files.
• Document processing and synthesis: Scan, process, and understand basic or complicated documents. Claude Sonnet 4 can even process hundreds of related documents without losing track of the overall context.
• Persistent AI agents: Develop your own AI agents that maintain persistence across workflows, tools calls, and more.

Real-world customers’ use cases

Although it was just released in beta, Anthropic customers are already taking advantage of Claude Sonnet 4’s expanded capacity.

• The team with Bolt.new uses Claude in their web-based application development platform of the same name.
• Headquartered in London, the team at iGent AI uses the latest version of Claude Sonnet 4 to convert normal conversations into usable, executable code.

While these use cases would technically be possible with previous iterations of Claude, the latest update allows for larger projects, streamlined workloads, and production-scale engineering.

Want to see where Claude is headed next? Don’t miss its new conversational voice mode in our TechRepublic coverage.