gpt · 2 arms · 4 effortsgpt-5.4-mini

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
medium
high
api
gpt-5.4-mini api noneopen run ↗
33.4K tokens8 runs · all valid
gpt-5.4-mini api lowopen run ↗
35.7K tokens8 runs · all valid
gpt-5.4-mini api mediumopen run ↗
157.6K tokens8 runs · all valid
gpt-5.4-mini api highopen run ↗
711.1K tokens8 runs · all valid
codex-oauth
—
gpt-5.4-mini codex-oauth lowopen run ↗
44.5K tokens4 runs · all valid
gpt-5.4-mini codex-oauth mediumopen run ↗
43.6K tokens4 runs · all valid
gpt-5.4-mini codex-oauth highopen run ↗
105.7K 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.

21.3×none → high output
0500K1Mnonelowmediumhigh711.1K33.4K
APIcodex (OAuth)
gpt-5.4-mini — model · weather-card-benchmark