Claude Sonnet 5.5 Released: Features, Price, Coding Performance and How to Use

Anthropic has officially released Claude Sonnet 5.5, the latest model in its Claude 5.5 AI family. The new model was introduced on September 28, 2026, as a faster and more cost-efficient option for everyday AI tasks, coding, document creation and longer workflows.

According to Anthropic, Claude Sonnet 5.5 generates responses more than 30% faster than Sonnet 5 and can cost up to 30% less per task because it generally requires fewer tokens to complete the same work.

What Is Claude Sonnet 5.5?

Claude Sonnet 5.5 is Anthropic's newest Sonnet model and the second model released in the Claude 5.5 generation.

Anthropic positions Sonnet 5.5 between its more powerful Opus model and lower-cost models for high-volume applications. It is designed particularly for well-defined everyday tasks, software development, bug fixing, documents, presentations, spreadsheets and rapid AI-assisted workflows.

The company says Sonnet 5.5 also improves image understanding and long-running tasks compared with the previous Sonnet generation.

Claude Sonnet 5.5 Key Features

The new AI model brings several important improvements.

30%+ Faster Output

Anthropic says Sonnet 5.5 generates outputs more than 30% faster than Sonnet 5.

This can be useful for developers, writers and professionals who repeatedly interact with AI while working on projects.

Improved Coding Performance

Coding is one of the major areas highlighted in the launch.

Anthropic reports that Sonnet 5.5 achieved 70.6% on Terminal-Bench 4.0, compared with 10.3% for Sonnet 5 in the company's reported evaluation. The benchmark measures agentic coding performance.

The model is designed for tasks such as fixing bugs, modifying existing code, building applications and working through multi-step development tasks.

Lower Cost Per Task

Sonnet 5.5 keeps the same standard API token pricing as Sonnet 5, but Anthropic says the new model can require fewer tokens to complete work.

The company estimates that Sonnet 5.5 can cost up to 30% less per task than its predecessor.

Better Documents and Office Work

Claude Sonnet 5.5 is also designed for producing polished documents, presentations and spreadsheets.

This makes the model relevant not only to programmers but also to students, researchers, businesses and professionals who use AI for knowledge work.

Claude Sonnet 5.5 Price

Anthropic lists the following API pricing for Claude Sonnet 5.5:

Input: $2 per 1 million tokens

Output: $10 per 1 million tokens

Cache reads: $0.20 per 1 million tokens

Cache writes: $2.50 per 1 million tokens

These are API prices rather than a general statement that every Claude user will pay these amounts. Access and subscription pricing can vary depending on the Claude product or cloud platform being used.

Claude Sonnet 5.5 vs Sonnet 5

Claude Sonnet 5.5 is positioned as a direct upgrade over Sonnet 5.

The biggest changes highlighted by Anthropic include faster generation, improved coding results, better efficiency and stronger performance on several knowledge-work evaluations.

Anthropic reports a Terminal-Bench 4.0 score of 70.6% for Sonnet 5.5 compared with 10.3% for Sonnet 5. However, benchmark results come from specific evaluation setups and should not be treated as a guarantee of performance for every real-world task.

Can Claude Sonnet 5.5 Be Used for Coding?

Yes. Coding is one of the main use cases for Claude Sonnet 5.5.

Developers can use the model to:

• Generate and modify code

• Find programming errors

• Explain existing code

• Debug applications

• Work on multi-step software projects

• Create documents and technical content

• Assist with development workflows

Anthropic says Sonnet 5.5 is available through the Claude Platform and also through major cloud platforms including Amazon Web Services, Google Cloud and Microsoft Azure.

Claude Sonnet 5.5 for Students and Content Creators

Students can use an AI model such as Sonnet 5.5 for research assistance, explanations, brainstorming and writing support, while content creators can use it for scripts, articles, outlines and editing.

However, AI-generated information should still be checked, especially when an article contains current news, technical specifications, statistics or other factual claims.

How to Use Claude Sonnet 5.5

Users and developers can access Claude Sonnet 5.5 through Anthropic's Claude ecosystem and supported cloud platforms.

For developers using the API, Anthropic identifies the model as:

claude-sonnet-5-5

The company says the model is available across the Claude Platform, Amazon Web Services, Google Cloud and Microsoft Azure.

Developers should check the current Claude documentation for account requirements, usage limits and the latest availability before starting a new project.

Claude Sonnet 5.5 Safety Features

Anthropic has also introduced additional safeguards with Sonnet 5.5.

The company says the model has cybersecurity safeguards because of its improved cyber capabilities. Anthropic also says higher-risk cybersecurity requests can fall back to Sonnet 5 under its safety system.

The company has additionally introduced measures intended to reduce large-scale attempts to extract model reasoning and capabilities.

Why Claude Sonnet 5.5 Is Getting Attention

The launch comes as AI companies continue developing models that can handle more complex software development and knowledge-work tasks.

Sonnet 5.5 is notable because Anthropic is targeting a combination of speed, performance and lower task-level cost rather than positioning it only as a maximum-capability model.

Anthropic's own testing reports strong improvements over Sonnet 5, while also noting that Opus 5.5 remains stronger for some complex, open-ended tasks requiring sustained judgment.

Conclusion

Claude Sonnet 5.5 is Anthropic's latest Sonnet AI model, released on September 28, 2026. It focuses on faster responses, coding, everyday knowledge work, document creation and improved efficiency.

Anthropic says the model is more than 30% faster than Sonnet 5 and can cost up to 30% less per task. Its reported Terminal-Bench 4.0 score also shows a substantial improvement over the previous Sonnet model in Anthropic's testing.

With API pricing of $2 per million input tokens and $10 per million output tokens, Sonnet 5.5 is positioned as a relatively cost-conscious option for developers and businesses that need capable AI for repeated workflows.