Thinking Levels · OpenAI
Last updated , from board data generated at 17:52 UTC.
This page compares GPT-5.6 Terra 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 Terra ranks #16 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 moves GPT-5.6 Terra only a little. Ultra lands 71 Elo above the run with no effort flag (medium), and the two ranges just overlap, though overlap alone doesn't test their difference. The bill moves a lot more: 4.4 times the cost, $0.275 a script, and about 8.0 minutes of waiting instead of 1.4.
Keep the flag off. On GPT-5.6 Terra, the effort setting changes your bill more than your script. Ultra does land closer to the target length and trims a bit of slop, but the overall Elo gain stays small. For new work, I'd skip this Terra. GPT-6 Sol (medium) ranks far above both settings for $0.056 a script, less than Terra with no flag, and comes back in about 0.8 minutes. If you still want a Terra, the newer GPT-6.1 Terra (low) lands in a range that overlaps the run with no flag here, for $0.027 a script and about 0.5 minutes.
Each point is one effort setting of GPT-5.6 Terra, 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) | 1922 ±39 | - | 80.7 | $0.062 | 1.4 min | #40 |
| ultra | 1993 ±42 | +71 | 81.4 | $0.275 | 8.0 min | #32 |
The 95% Writing Elo ranges overlap: GPT-5.6 Terra (medium) runs from 1879 to 1957, and GPT-5.6 Terra (ultra) from 1952 to 2037.
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.