Claude Sonnet 4.6
The balanced middle. Strong reasoning without the Opus price.
A model a workspace can run · gbrain.io
Anthropic
Claude Sonnet 4.6
The balanced middle. Strong reasoning without the Opus price.
- Vendor
- Anthropic
- Released
- May 2026
- Model id
- claude-sonnet-4-6
Sonnet sits between Haiku and Opus: noticeably stronger than Haiku on reasoning and coding, noticeably cheaper than Opus. For a lot of everyday agent work it’s the sweet spot: capable enough to trust, light enough to run often.
Best for: the default when Opus is overkill but Haiku isn’t enough.
Pick it per workspace from the model settings and switch any time. Inference is included in the plan, or bring a key of the team's own.