AISystem

Claude Sonnet 4.6

Anthropic's mid-tier model of February 2026, which took the lead on GDPval-AA and Finance Agent v1.1 ahead of its own larger sibling Opus 4.6 at a fraction of the cost.

reasoning-model

Sources

Attribution and reporting linked by this record; links do not imply independent verification.

Claude Sonnet 4.6 is the smaller, cheaper tier of Anthropic’s Claude 4.6 generation, released after the larger Claude Opus 4.6. In this corpus it enters as the cheap model that beat the expensive one, claiming 1633 Elo on GDPval-AA and 63.3% on Finance Agent v1.1, a datapoint for the deflating price of reasoning.

Development timeline

Newest reporting dates first. Reporting dates may differ from event dates.

    Knowledge graphopen in browser →

    Same as
    Developed by
    Evaluated on

    Cite or correct this record

    Suggest a correction — opens a GitHub draft for you to review and submit.