Thinking Levels · OpenAI
GPT-6.1 Terra: Writing Quality by Reasoning Effort
Last updated , from board data generated at 20:15 UTC.
This page compares GPT-6.1 Terra at 5 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-6.1 Terra ranks #23 of 84 models through its best setting. The other ranks on this page are configuration ranks: one setting's place among all 141 configurations.
What More Effort Changes for GPT-6.1 Terra
Effort barely moves the Writing Elo rating on GPT-6.1 Terra. All 5 settings land within 25 Elo of each other, their ranges overlap, and they don't even line up in effort order. From low to max, the gain is 3 Elo, so max beats low in about 50% of script comparisons, a coin flip. Below max, the setting barely changes what the model does: low through xhigh spend next to no reasoning tokens on a draft, while max spends thousands. Low through xhigh also cost and take the same, about $0.027 and 0.5 minutes a script. Max costs 3.0 times as much and takes about 1.8 minutes.
Recommended Setting
Use low. It's the cheapest and fastest setting, and no setting above it gains much Writing Elo. Max does land closer to the target length and writes better visual cues, but it gives a little voice and flow back and ends up close to low on Writing Elo, for more money and a longer wait. I'd skip it.
Next to the previous Terra, it doesn't write better. GPT-5.6 Terra (ultra) ranks above all 5 settings here, with a range clear of each one, though it costs $0.275 and takes about 8.0 minutes a script. GPT-5.6 Terra (medium) lands in a range that overlaps low's, for 2.3 times the price and about 1.4 minutes a script. So this is the cheap, fast Terra, good for first drafts in bulk. When the script matters, GPT-6 Sol (medium) ranks far higher for $0.056 a script.
Every Setting on the Chart
Each point is one effort setting of GPT-6.1 Terra, joined from the lowest setting to the highest. Higher and further left is better: a higher score for less money or less time.
How to Use the Chart
- A hollow marker is the provider default: the level the provider applies when the request sets no effort level.
- The tabs choose the score on the vertical axis, X axis switches between cost and time, and Table lists the same numbers as rows.
- Select a point for its numbers and a link to its scorecard.
On a small screen, Table lists the same numbers as rows.
All Effort Settings, Step by Step
| Setting | Writing Elo | Step | Writing Score | Cost / Script | Time / Script | Configuration Rank |
|---|---|---|---|---|---|---|
| low | 1920 ±41 | - | 80.2 | $0.027 | 0.5 min | #46 |
| medium | 1898 ±45 | -22 | 79.9 | $0.027 | 0.5 min | #49 |
| high | 1922 ±38 | +24 | 80.3 | $0.027 | 0.5 min | #44 |
| xhigh | 1906 ±37 | -16 | 80.1 | $0.028 | 0.5 min | #48 |
| max | 1923 ±38 | +17 | 80.3 | $0.079 | 1.8 min | #43 |
The 95% Writing Elo ranges overlap: GPT-6.1 Terra (low) runs from 1879 to 1961, and GPT-6.1 Terra (max) from 1884 to 1959.
Other OpenAI Models by Effort
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.