gpt · 2 arms · 4 effortsgpt-5.3-codex

One model across every effort and arm. Each cell is a run — open it for the full card. Failure chips mark runs with no valid slot.

effort strip

arm \ effort
none
low
high
xhigh
cursor
gpt-5.3-codex cursor noneopen run ↗
17.9K tokens4 runs · all valid
gpt-5.3-codex cursor lowopen run ↗
15.2K tokens4 runs · all valid
gpt-5.3-codex cursor highopen run ↗
25.6K tokens4 runs · all valid
gpt-5.3-codex cursor xhighopen run ↗
56K tokens4 runs · all valid
cursor-xl
gpt-5.3-codex cursor-xl noneopen run ↗
15.5K tokens4 runs · all valid
gpt-5.3-codex cursor-xl lowopen run ↗
15.1K tokens4 runs · all valid
gpt-5.3-codex cursor-xl highopen run ↗
23.6K tokens4 runs · all valid
gpt-5.3-codex cursor-xl xhighopen run ↗
53K tokens4 runs · all valid

effort → output size

Output completion tokens per effort, per arm. A read on how much each model spends as effort rises — not a quality axis.

3.7×low → xhigh output
050K100Knonelowhighxhigh56K15.2K
cursorcursor-xl
gpt-5.3-codex — model · weather-card-benchmark