Claude Sonnet
AnthropicClaude Sonnet is Anthropic's mid-tier Claude model, positioned between Haiku (small, high-volume) and Opus (complex, judgment-heavy work), and aimed at everyday coding, agent and knowledge-work tasks. The name is applied to whichever generation is currently shipping rather than to one fixed release, from Sonnet 4.6 through Sonnet 5 to Sonnet 5.5. Weights are closed, and the model is available through the Anthropic API, cloud platforms and the Claude apps.
Openness
1 high confidence- weights
- closed
- data
- closed
- code
- closed
- license
- Proprietary(API-only)
Claude Sonnet is sold only through Anthropic's API, the Claude apps and cloud platforms, and no release in the line, including the newest, has shipped weights, training data or training code.
- https://www.anthropic.com/claude-sonnet-5-5 recorded 2026-09-29
the tier current release, Claude Sonnet 5.5 (September 28, 2026): "Claude Sonnet 5.5 is now available on all platforms, including Amazon Web Services, Google Cloud, and Microsoft Azure". Searched the body for weights, license, open source, open-weight and self-host: none; no weights, corpus or training code offered.
- https://www.anthropic.com/news/claude-sonnet-5 recorded 2026-09-29
the prior release, Claude Sonnet 5 (June 30, 2026): "Developers can use" claude-sonnet-5 "via the" Claude API, priced at "$2 per million input tokens and $10 per million output tokens". Searched the body for weights, license, open source, open-weight and self-host: none.
Adoption
5 medium confidenceSonnet is offered on every Claude plan, the API and the major cloud platforms, and at the Sonnet 5 release it became the default model for the Free and Pro plans. Anthropic publishes no per-model usage figure, so this reading is a judgment from reported traction rather than a measured count.
- https://www.anthropic.com/claude-sonnet-5-5 recorded 2026-09-29
Sonnet 5.5 page: "Claude Sonnet 5.5 is now available on all platforms, including Amazon Web Services, Google Cloud, and Microsoft Azure"; no usage or user figure published
- https://www.anthropic.com/news/claude-sonnet-5 recorded 2026-09-29
the prior release: "it is the default model for Free and Pro plans, and is available to Max, Team, and Enterprise users"; no usage or user figure published
Capability
4 medium confidenceAnthropic reports that Sonnet 5.5 scores 70.6% on Terminal-Bench 4.0 and, at maximum effort, performs comparably to Opus 5.5 on several evaluations. Anthropic also states that Opus 5.5 remains clearly stronger at complex, open-ended work, which keeps Sonnet a step below Opus.
- https://www.anthropic.com/claude-sonnet-5-5 recorded 2026-09-29
Sonnet 5.5 (September 28, 2026): "Sonnet 5.5 scores 70.6% on Terminal-Bench 4.0"; "Opus 5.5 remains clearly stronger at complex, open-ended work requiring sustained judgment". No SWE-bench Verified figure appears.
Verified 2026-09-29