OpenAI DevDay 2026 AI

Theo Browne's post on X, curated on eChai

  • Models and speed

    Curated from X →

    Theo Browne's second run, which found 6.1 Sol doing far better in Codex than in mini-swe, the harness Artificial Analysis uses.

    Theo Browne Theo Browne

    Updated run just finished. GPT-6.1 Sol still crushes. Turns out it performs WAY better in Codex than in mini-swe (what Artificial Analysis uses)

eChai Partner Brands The partnership →