Side-by-side benchmark comparison across knowledge, coding, math, and reasoning.
Gemini 3.1 Pro is clearly ahead on the aggregate, 89 to 56. The gap is large enough that you do not need to squint at the spreadsheet to see the difference.
Gemini 3.1 Pro's sharpest advantage is in coding, where it averages 79 against 49.3. The single biggest benchmark swing on the page is SWE-bench Verified, 75 to 49.3.
Gemini 3.1 Pro is also the more expensive model on tokens at $1.25 input / $5.00 output per 1M tokens, versus $1.10 input / $4.40 output per 1M tokens for o3-mini. o3-mini is the reasoning model in the pair, while Gemini 3.1 Pro is not. That usually helps on harder chain-of-thought-heavy tests, but it can also mean more latency and more token spend in real use. Gemini 3.1 Pro gives you the larger context window at 1M, compared with 200K for o3-mini.
Pick Gemini 3.1 Pro if you want the stronger benchmark profile. o3-mini only becomes the better choice if you want the cheaper token bill or you want the stronger reasoning-first profile.
Gemini 3.1 Pro
86
o3-mini
82.1
Gemini 3.1 Pro
79
o3-mini
49.3
Gemini 3.1 Pro
97.1
o3-mini
87.3
Gemini 3.1 Pro
95
o3-mini
93.9
Gemini 3.1 Pro is ahead overall, 89 to 56. The biggest single separator in this matchup is SWE-bench Verified, where the scores are 75 and 49.3.
Gemini 3.1 Pro has the edge for knowledge tasks in this comparison, averaging 86 versus 82.1. Inside this category, GPQA is the benchmark that creates the most daylight between them.
Gemini 3.1 Pro has the edge for coding in this comparison, averaging 79 versus 49.3. Inside this category, SWE-bench Verified is the benchmark that creates the most daylight between them.
Gemini 3.1 Pro has the edge for math in this comparison, averaging 97.1 versus 87.3. Inside this category, AIME 2024 is the benchmark that creates the most daylight between them.
Gemini 3.1 Pro has the edge for instruction following in this comparison, averaging 95 versus 93.9. Inside this category, IFEval is the benchmark that creates the most daylight between them.
Get notified when new models drop, benchmark scores change, or the leaderboard shifts. One email per week.
Free. No spam. Unsubscribe anytime. We only store derived location metadata for consent routing.