Thinking Levels · OpenAI
Last updated , from board data generated at 17:52 UTC.
This page compares GPT-5.6 Luna at 2 reasoning-effort settings, each tested on the same 10 scripts: the Writing Score (the 0 to 100 consensus of the judges and code), the Writing Elo (a rating from pairwise comparisons of those scores), and the cost and time of each setting. As a model, GPT-5.6 Luna ranks #30 of 87 models through its best setting. The other ranks on this page are configuration ranks: one setting's place among all 151 configurations.
Ultra effort lifts GPT-5.6 Luna a long way, mostly by fixing length. Ultra takes it from #85 to #65, 198 Elo above the run with no effort flag (medium), and the ranges sit far apart. With no flag, Luna keeps missing the target length, and at ultra it mostly lands it. The script still costs pennies at $0.026. The wait is longer, though: about 3.1 minutes a script instead of 0.5.
Run it at ultra. The gain is large, the ranges are clearly separated, and the extra cost is tiny in absolute terms. Keep expectations in check, though: even at its best, GPT-5.6 Luna sits at #65, so treat it as a cheap drafting model and plan an editing pass.
Worth knowing: GPT-5.6 Luna (ultra) ranks above every setting of the newer GPT-6 Luna, though its range overlaps that of GPT-6 Luna (max). For about the same price, though, GPT-6.1 Terra (low) ranks clearly above it and comes back in about 0.5 minutes, so among the really cheap OpenAI options, that's my pick now. If you can spend a little more, GPT-6 Sol (medium) ranks far higher for $0.056 a script.
Each point is one effort setting of GPT-5.6 Luna, joined from the lowest setting to the highest. Higher and further left is better: a higher score for less money or less time.
On a small screen, Table lists the same numbers as rows.
| Setting | Writing Elo | Step | Writing Score | Cost / Script | Time / Script | Configuration Rank |
|---|---|---|---|---|---|---|
| medium (provider default) | 1533 ±41 | - | 76.7 | $0.0065 | 0.5 min | #85 |
| ultra | 1731 ±41 | +198 | 78.8 | $0.026 | 3.1 min | #65 |
Writing Elo comes from comparing every pair of models on every script, which puts these settings on the same scale as the leaderboard. Step is the Writing Elo change from the provider default. Cost is one full script at API list prices, and time is the average for one accepted script, retries included. The thinking-levels page covers the other models, and the OpenAI family page lists every OpenAI model side by side.