Compare · ModelsLive · 2 picked · head to head
Gemini 3.1 Flash Lite vs o1
Side by side · benchmarks, pricing, and signals you can act on.
Winner summary
Gemini 3.1 Flash Lite wins on 8/8 benchmarks
Gemini 3.1 Flash Lite wins 8 of 8 shared benchmarks. Leads in knowledge · general · math.
Category leads
knowledge·Gemini 3.1 Flash Litegeneral·Gemini 3.1 Flash Litemath·Gemini 3.1 Flash Litecoding·Gemini 3.1 Flash Lite
Hype vs Reality
Attention vs performance
Gemini 3.1 Flash Lite
#162 by perf·#5 by attention
o1
#158 by perf·no signal
Best value
Gemini 3.1 Flash Lite
42.0x better value than o1
Gemini 3.1 Flash Lite
50.9 pts/$
$0.88/M
o1
1.2 pts/$
$37.50/M
Vendor risk
Who is behind the model
Google DeepMind
$4.20T·Tier 1
OpenAI
$840.0B·Tier 1
Head to head
8 benchmarks · 2 models
Gemini 3.1 Flash Liteo1
Chess Puzzles
Gemini 3.1 Flash Lite leads by +10.5
Chess Puzzles · tests strategic and tactical reasoning by having models solve chess puzzle positions, evaluating lookahead and pattern recognition abilities.
Gemini 3.1 Flash Lite
21.1
o1
10.6
Dtbench
Gemini 3.1 Flash Lite leads by +3.5
Gemini 3.1 Flash Lite
61.3
o1
57.8
FrontierMath-Tiers-1-3-v2-Private
Gemini 3.1 Flash Lite leads by +13.0
Gemini 3.1 Flash Lite
27.7
o1
14.7
GPQA diamond
Gemini 3.1 Flash Lite leads by +6.7
Graduate-Level Google-Proof QA (Diamond set) · expert-crafted questions in physics, biology, and chemistry that are difficult even for domain PhDs.
Gemini 3.1 Flash Lite
75.8
o1
69.0
HLE
Gemini 3.1 Flash Lite leads by +0.7
HLE (Humanity's Last Exam) · a reasoning benchmark designed to be the hardest public evaluation of AI. Questions span mathematics, physics, philosophy, and logic · curated to be at or beyond the frontier of human expert capability. Tested with and without tool augmentation. Claude Opus 4.7 scores 46.9% without tools and 54.7% with tools · making it one of the few benchmarks where the top score is below 60%.
Gemini 3.1 Flash Lite
4.0
o1
3.3
Lmca
Gemini 3.1 Flash Lite leads by +15.0
Gemini 3.1 Flash Lite
41.2
o1
26.2
OTIS Mock AIME 2024-2025
Gemini 3.1 Flash Lite leads by +6.7
OTIS Mock AIME 2024-2025 · simulated American Invitational Mathematics Examination problems testing advanced problem-solving skills.
Gemini 3.1 Flash Lite
80.0
o1
73.3
WeirdML
Gemini 3.1 Flash Lite leads by +6.1
WeirdML · tests models on unusual and adversarial machine learning tasks that require creative problem-solving beyond standard patterns.
Gemini 3.1 Flash Lite
52.2
o1
46.1
Full benchmark table
| Benchmark | Gemini 3.1 Flash Lite | o1 |
|---|---|---|
Chess Puzzles Chess Puzzles · tests strategic and tactical reasoning by having models solve chess puzzle positions, evaluating lookahead and pattern recognition abilities. | 21.1 | 10.6 |
Dtbench | 61.3 | 57.8 |
FrontierMath-Tiers-1-3-v2-Private | 27.7 | 14.7 |
GPQA diamond Graduate-Level Google-Proof QA (Diamond set) · expert-crafted questions in physics, biology, and chemistry that are difficult even for domain PhDs. | 75.8 | 69.0 |
HLE HLE (Humanity's Last Exam) · a reasoning benchmark designed to be the hardest public evaluation of AI. Questions span mathematics, physics, philosophy, and logic · curated to be at or beyond the frontier of human expert capability. Tested with and without tool augmentation. Claude Opus 4.7 scores 46.9% without tools and 54.7% with tools · making it one of the few benchmarks where the top score is below 60%. | 4.0 | 3.3 |
Lmca | 41.2 | 26.2 |
OTIS Mock AIME 2024-2025 OTIS Mock AIME 2024-2025 · simulated American Invitational Mathematics Examination problems testing advanced problem-solving skills. | 80.0 | 73.3 |
WeirdML WeirdML · tests models on unusual and adversarial machine learning tasks that require creative problem-solving beyond standard patterns. | 52.2 | 46.1 |
Pricing · per 1M tokens · projected $/mo at 10M tokens
| Model | Input | Output | Context | Projected $/mo |
|---|---|---|---|---|
| $0.25 | $1.50 | 1.0M tokens (~524 books) | $5.63 | |
| $15.00 | $60.00 | 200K tokens (~100 books) | $262.50 |