Noometry

Google, proprietary

Gemini 3.1 Pro Preview

Gemini 3.1 Pro Preview by Google ranks 23rd of 354 ranked models on the Noometry Index as of October 2026, with a score of 56.7. Its strongest category is knowledge, where it ranks 3rd. API pricing starts at $2 per million input tokens and $12 per million output tokens, with a 1.05M-token context window.

Last verified

Noometry rank
#23 of 354
Index score
56.7
Evidence
Confirmed 71 results
Provider
Google
Released
February 19, 2026
Weights
Proprietary
Reasoning
Yes
Context window
1.05M
Max output
66K
Input price
$2 / M
Output price
$12 / M
Blended price
$4.50 / M
Output speed
Not measured
Value
#175 of 219
Knowledge cutoff
January 2025
Input
text, image, video, audio, pdf

Category scores

Each category score combines every public result we have in that category.

  1. Coding 42.5
  2. Agentic & Tool Use 37.7
  3. Reasoning 71.7
  4. Math 62.1
  5. Knowledge 71.8
  6. Multimodal 37.9
  7. Multilingual 57.0
  8. Instruction Following 77.0
  9. Long Context 47.4
  10. Writing & Preference 66.1
CategoryScoreRankResults
Coding42.5#998
Agentic & Tool Use37.7#349
Reasoning71.7#1213
Math62.1#346
Knowledge71.8#35
Multimodal37.9#693
Multilingual57.0#121
Instruction Following77.0#321
Long Context47.4#183
Writing & Preference66.1#375

Strengths and weaknesses

Categories where Gemini 3.1 Pro Preview places highest and lowest among the models ranked in each, with its score against that category's median.

Strongest categories

CategoryScorevs medianRank
Knowledge71.8+34.5#3 of 314, top 1%
Reasoning71.7+48.1#12 of 350, top 4%
Multilingual57.0+9.6#12 of 297, top 5%

Weakest categories

CategoryScorevs medianRank
Multimodal37.9−0.7#69 of 128, top 54%
Coding42.5+3.8#99 of 340, top 30%
Agentic & Tool Use37.7+7.4#34 of 154, top 23%

Closest competitors

The models ranked just above and below Gemini 3.1 Pro Preview. When scores are this close, price and speed are often the better way to choose.

ModelRankScoreBlended $/MSpeed
Claude Opus 4.7#1958.3$1033Compare
Claude Opus 4.6#2058.2$1019Compare
Grok 4.6#2156.9$3—Compare
Qwen3.8 Max#2256.8$3—Compare
Gemini 4 Argon#2456.5——Compare
Grok 4.5#2555.0$34Compare
GLM-5.3#2654.8$2.15—Compare
Muse Spark 1.3#2754.8$2—Compare

Sponsored placements are available on pages like this one. Advertise on Noometry

Benchmark results

Every published result we track, with its source. Bold rows are the ones used for ranking; where several exist we prefer independent runs over self-reported numbers.

Coding

BenchmarkScorePositionSettingSourceDate
SWE-bench Verified75.6%#12 of 32, top 38%Epoch AI2026-02-24
DeepSWE11.7%#29 of 29, top 100%Epoch AI
LMArena WebDev1447#56 of 113, top 50%LMArena2026-10-08
SciCode58.9%#11 of 121, top 10%Epoch AI
GSO22.6%#13 of 31, top 42%Epoch AI
WeirdML72.1%#17 of 119, top 15%Epoch AI
LMArena Coding1484#39 of 294, top 14%LMArena2026-10-08
MirrorCode8.9%#9 of 9, top 100%highEpoch AI2026-08-10
ALE-Bench1,161#35 of 105, top 34%Epoch AI
AlgoTune2.02#2 of 18, top 12%Epoch AI

Agentic & Tool Use

BenchmarkScorePositionSettingSourceDate
Terminal-Bench80.2%#4 of 41, top 10%Epoch AI
APEX-Agents35.3%#41 of 49, top 84%Epoch AI
τ²-bench Banking26%#16 of 26, top 62%highτ²-bench2026-05-05
DeepResearch Bench47.8%#10 of 24, top 42%highEpoch AI
DeepResearch Bench44.5%lowEpoch AI
PostTrainBench22%#10 of 11, top 91%Epoch AI
BALROG57%#5 of 35, top 15%Epoch AI
ExploitBench26.1%#4 of 9, top 45%Epoch AI
GBAEval0.8%#19 of 23, top 83%Epoch AI
GDP.pdf17%#24 of 36, top 67%Epoch AI
GDP.pdf17%highEpoch AI
LMArena Search1211#9 of 32, top 29%LMArena2026-08-24
METR Time Horizons77%#3 of 32, top 10%Epoch AI
Vending-Bench 2911.21Epoch AI
Vending-Bench 23,774#38 of 60, top 64%Epoch AI

Reasoning

BenchmarkScorePositionSettingSourceDate
ARC-AGI-277.1%#16 of 83, top 20%Epoch AI
SimpleBench79.6%#3 of 77, top 4%Epoch AI
NYT Connections (extended)97.4%#2 of 91, top 3%Lech Mazur benchmarks
ARC-AGI-198%#6 of 83, top 8%Epoch AI
CritPt17.7%#31 of 134, top 24%Epoch AI
Chess Puzzles55%#7 of 129, top 6%Epoch AI2026-02-19
Chess Puzzles49%highEpoch AI2026-08-06
EnigmaEval36.8%#3 of 38, top 8%Epoch AI
Thematic Generalization79.4%#3 of 23, top 14%Lech Mazur benchmarks
EBR-Bench14.3%#16 of 24, top 67%Epoch AI2026-06-25
LMArena Hard Prompts1485#22 of 297, top 8%LMArena2026-10-08
Mystery Game Puzzles34%#21 of 74, top 29%highEpoch AI2026-07-27
Mystery Game Puzzles29%lowEpoch AI2026-08-06
Mystery Game Puzzles32%mediumEpoch AI2026-08-06
DTBench95.5%highEpoch AI
DTBench94.9%lowEpoch AI
DTBench97.1%#8 of 151, top 6%mediumEpoch AI
LMCA53.8%#15 of 125, top 12%highEpoch AI
LMCA51.2%lowEpoch AI
LMCA52.8%mediumEpoch AI
Epoch Capabilities Index154.77#33 of 213, top 16%Epoch AI2026-02-19
ForecastBench59#43 of 72, top 60%Epoch AI
ForecastBench58.1highEpoch AI

Math

BenchmarkScorePositionSettingSourceDate
FrontierMath (Tiers 1-3)59.6%#35 of 81, top 44%Epoch AI2026-06-11
FrontierMath Tier 426.8%#35 of 63, top 56%Epoch AI2026-06-11
MathArena Final-Answer Competitions86.5%#4 of 29, top 14%MathArena
OTIS Mock AIME 2024-202595.6%#31 of 173, top 18%Epoch AI2026-02-20
OTIS Mock AIME 2024-202595.6%highEpoch AI2026-08-06
ProofBench26%#43 of 77, top 56%Epoch AI
LMArena Math1485#24 of 285, top 9%LMArena2026-10-08
FrontierMath (Feb 2025 set)36.9%#13 of 68, top 20%Epoch AI2026-02-19
FrontierMath Tier 4 (v1)16.7%#12 of 55, top 22%Epoch AI2026-02-19

Knowledge

BenchmarkScorePositionSettingSourceDate
GPQA Diamond94.1%Epoch AI2026-02-20
GPQA Diamond94.4%#7 of 186, top 4%highEpoch AI2026-08-06
Humanity's Last Exam46.4%#3 of 41, top 8%Epoch AI
SimpleQA Verified73.5%#3 of 77, top 4%highEpoch AI2026-08-10
Vectara Hallucination Rate (lower is better)10.4%#57 of 96, top 60%Vectara Hallucination Leaderboard
LMArena Expert1485#37 of 273, top 14%LMArena2026-10-08

Multimodal

BenchmarkScorePositionSettingSourceDate
LMArena Vision1296#18 of 122, top 15%LMArena2026-10-09
Blueprint-Bench 226.5%#20 of 31, top 65%Epoch AI
Furniture Assembly26.7%#25 of 31, top 81%highEpoch AI2026-09-10
LMArena Document1444#26 of 38, top 69%LMArena2026-09-13

Multilingual

BenchmarkScorePositionSettingSourceDate
LMArena Non-English1477#12 of 297, top 5%LMArena2026-10-08
LMArena Chinese1529#17 of 285, top 6%LMArena2026-10-08
LMArena French1487#23 of 223, top 11%LMArena2026-10-08
LMArena German1491#13 of 231, top 6%LMArena2026-10-08
LMArena Japanese1493#9 of 211, top 5%LMArena2026-10-08
LMArena Korean1455#15 of 213, top 8%LMArena2026-10-08
LMArena Russian1498#8 of 283, top 3%LMArena2026-10-08
LMArena Spanish1479#14 of 226, top 7%LMArena2026-10-08

Instruction Following

BenchmarkScorePositionSettingSourceDate
LMArena Instruction Following1466#28 of 298, top 10%LMArena2026-10-08

Long Context

BenchmarkScorePositionSettingSourceDate
CL-bench20.8%#5 of 19, top 27%Epoch AI
CL-bench Life16.9%#5 of 13, top 39%Epoch AI
LMArena Longer Query1483#17 of 291, top 6%LMArena2026-10-08

Writing & Preference

BenchmarkScorePositionSettingSourceDate
LMArena Text1481#15 of 297, top 6%LMArena2026-10-08
LMArena Creative Writing1482#11 of 295, top 4%LMArena2026-10-08
EQ-Bench Creative Writing1491#58 of 115, top 51%EQ-Bench
EQ-Bench 41142#21 of 28, top 75%EQ-Bench
LMArena Multi-Turn1488#12 of 295, top 5%LMArena2026-10-08

API pricing by provider

RouteInput $/MOutput $/MCached input $/MChecked
google$2$12$0.202026-10-10
openrouter$2$12$0.202026-10-10
vertex$2$12$0.202026-10-10

All Google API prices → Estimate your cost →

Compare Gemini 3.1 Pro Preview

Other Google models

Frequently asked questions

How good is Gemini 3.1 Pro Preview?

Gemini 3.1 Pro Preview by Google ranks 23rd of 354 ranked models on the Noometry Index as of October 2026, with a score of 56.7. Its strongest category is knowledge, where it ranks 3rd. API pricing starts at $2 per million input tokens and $12 per million output tokens, with a 1.05M-token context window.

How much does Gemini 3.1 Pro Preview cost?

Gemini 3.1 Pro Preview costs $2 per million input tokens and $12 per million output tokens on Google's own API, with cached input at $0.20.

What is Gemini 3.1 Pro Preview's context window?

Gemini 3.1 Pro Preview accepts up to 1.05M tokens of input and can write up to 66K tokens in one response.

Is Gemini 3.1 Pro Preview open source?

No. Gemini 3.1 Pro Preview is proprietary and available only through Google's API and partner platforms.

What are Gemini 3.1 Pro Preview's strengths and weaknesses?

Relative to other ranked models, Gemini 3.1 Pro Preview places best in knowledge, reasoning, multilingual and lowest in multimodal, coding, agentic & tool use.

What is Gemini 3.1 Pro Preview best at?

Its best category is knowledge, where it ranks 3rd on Noometry.