What is Claude Sonnet 5.5?
Claude Sonnet 5.5 is Anthropic's mid-tier Claude model, which Anthropic describes as its best combination of speed and intelligence, for well-scoped coding, bug fixes and documents such as slides and spreadsheets. Use it for daily coding tasks or a first draft of a report deck. It costs half of Opus 5.5 per token, but code written for Sonnet 5 can break on five API changes.
What can you do with it?
Anthropic positions Sonnet 5.5 as a faster, lower-cost complement to Claude Opus 5.5: Opus for complex work that needs careful judgment, Sonnet for well-scoped everyday tasks. A typical coding use is giving it a bug report and the relevant files and asking for a fix plus a test that proves it.
For documents, Anthropic says Sonnet 5.5 can follow a slide template to build a deck; in one internal test it produced a 10-slide operating review from earnings materials. In agent systems, Opus 5.5 can plan while Sonnet 5.5 implements. Anthropic also states that Opus 5.5 remains clearly stronger at complex, open-ended work. Benchmark figures in the announcement are Anthropic's own and were not tested by AppFor.ai.
Try it on a small task
Suggested workflow: take five closed bug tickets from your own repository where you know the accepted fix. Give Sonnet 5.5 each ticket and the relevant files, at the default high effort, and ask for a patch and a test.
Compare each patch with the fix your team merged: does it address the same cause, stay inside the needed files and pass your tests? Then rerun the failures at a higher effort or with Opus 5.5 and compare cost per accepted patch. In the Claude apps, you can run the same check on a short deck by attaching your template and source notes.
Pricing and access
Anthropic released Sonnet 5.5 on September 28, 2026, with the API ID claude-sonnet-5-5. API pricing is $2 per million input tokens and $10 per million output tokens, the same as Sonnet 5. Cache writes cost $2.50 for 5 minutes and $4 for 1 hour, cache reads $0.10, and the Batch API halves input and output prices. It is also offered on Amazon Bedrock, Claude Platform on AWS, Claude Models on Google Cloud and Microsoft Foundry.
On Claude's pricing page, Sonnet is listed for Free, Pro and Max plans; usage limits apply and exact per-plan limits are not established from the pages we checked. If you are migrating from Sonnet 5, thinking disabled now returns an error, so you send between_tools instead; forced tool use returns a 400 error; and the older computer_20251124 tool is not accepted on the Claude API and Google-managed Google Cloud.
What stands out
- 1M-token context window and 128K max output tokens
- Adaptive thinking with high as the default effort
- 300K output tokens through the Batch API (beta)
- Available on Claude API, Amazon Bedrock, Google Cloud and Microsoft Foundry
Before you choose
- Anthropic says Opus 5.5 stays clearly stronger on complex, open-ended work.
- Five breaking API changes when moving from Sonnet 5.
- Thinking blocks only work in the account that produced them.
Go to the source
Sources & further reading
Based on the official documentation below. Features and access can change; check the provider before making a decision.
Was this page useful?
Help us make our coverage more useful. This is feedback on the page, not a rating of the tool.
Go a little further