Thinking Levels · Google
Last updated , from board data generated at 17:52 UTC.
This page compares Gemini 3.5 Flash 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, Gemini 3.5 Flash ranks #60 of 87 models through its best setting. The other ranks on this page are configuration ranks: one setting's place among all 151 configurations.
Gemini 3.5 Flash's Writing Elo barely changes with high thinking. High and the medium thinking you get with no flag land 3 Elo apart, with ranges almost on top of each other. High still costs more, $0.136 a script against $0.111, and that's already a lot for a Flash model.
If you're on Gemini 3.5 Flash, keep medium thinking. High thinking barely moves Writing Elo. But I'd move off this model for scripts. Even with no flag, it costs about what Gemini 3.1 Pro (low thinking) does, and at #117 it sits near the bottom of Google's thinking ladders. Gemini 2.5 Pro (dynamic thinking) ranks above both settings for $0.067 a script. Its range is clear of both of theirs, if only just, and Gemini 3.5 Flash with no flag costs 1.7 times as much.
Each point is one effort setting of Gemini 3.5 Flash, 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 | Reasoning Tokens | Configuration Rank |
|---|---|---|---|---|---|---|---|
| medium thinking (provider default) | 1029 ±43 | - | 69.5 | $0.111 | 0.8 min | 7,065 | #117 |
| high | 1032 ±45 | +3 | 69.5 | $0.136 | 1.0 min | 9,818 | #116 |
The 95% Writing Elo ranges overlap: Gemini 3.5 Flash (medium thinking) runs from 988 to 1075, and Gemini 3.5 Flash (high thinking) from 983 to 1073.
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 Google family page lists every Google model side by side.