Thinking Levels · OpenAI
Last updated , from board data generated at 17:52 UTC.
This page compares GPT-5.5 at 4 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-5.5 ranks #41 of 87 models through its best setting. The other ranks on this page are configuration ranks: one setting's place among all 151 configurations.
More effort lifts GPT-5.5 a little, and on this model the extra thinking is almost free. From its lowest setting to xhigh, it climbs 109 Elo, so xhigh would beat the lowest setting in about 65% of script comparisons. The price per script barely moves, $0.174 at the bottom and $0.184 at the top, and neither does the wait.
On GPT-5.5, I'd set it to xhigh. It has the best score here, though its range still overlaps the one you get with no flag (medium), and each step overlaps the next. Overlap alone doesn't say which of those settings writes better.
The real problem is length, and thinking only fixes a slice of it. At every setting, GPT-5.5 misses the target length badly, far worse than the newer GPT models. Its voice score is fine, and about the same at every setting. Unless it's already wired into your setup, I'd move on. GPT-6 Sol (medium) ranks far above every setting here, lands much closer to length, and costs $0.056 a script, less than any of them.
Each point is one effort setting of GPT-5.5, 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 |
|---|---|---|---|---|---|---|
| none | 1421 ±54 | - | 75.2 | $0.174 | 0.8 min | #97 |
| medium (provider default) | 1460 ±46 | - | 75.8 | $0.176 | 1.0 min | #93 |
| high | 1516 ±47 | +95 | 76.4 | $0.183 | 0.9 min | #90 |
| xhigh | 1530 ±42 | +14 | 76.8 | $0.184 | 0.9 min | #87 |
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 previous explicit setting. 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.