Claude Sonnet 4.6

Hybrid reasoning with a 1M context window for agents and high-volume jobs.

Claude Sonnet 4.6 is the previous Sonnet card on the official page, dated February 17, 2026. Anthropic still describes it as a hybrid reasoning model with a 1M context window for real-time agents and high-volume work. If a team froze an eval on 4.6, this is the page I keep in the catalog so we do not pretend it vanished.

The Sonnet 5 launch is explicit: 5 is a substantial improvement over 4.6 on reasoning, tool use, coding, and knowledge work. I would not start a new agent on 4.6 unless a contract pins the id. For new work, read Claude Sonnet 5 first.

See also agents and all versions.

Open All models