Noometry

Model comparison

Gemini 3.1 Pro Preview vs Grok 4.20 (Non-Reasoning)

Gemini 3.1 Pro Preview is the stronger model overall, scoring 56.7 to 48.5 on the Noometry Index. Grok 4.20 (Non-Reasoning) costs 2.9× less per token, which makes it the better buy when Gemini 3.1 Pro Preview's lead doesn't matter for your workload.

Last verified . 45 shared benchmarks.

Gemini 3.1 Pro Preview Google

56.7

Rank #20 Confirmed

Grok 4.20 (Non-Reasoning) xAI

48.5

Rank #48 Confirmed

Summary

  • They share 45 benchmarks with published results for both. Gemini 3.1 Pro Preview scores higher in 10 categories and Grok 4.20 (Non-Reasoning) in 0 categories; 8 gaps are clear of the uncertainty.
  • The widest gap is in reasoning, where Gemini 3.1 Pro Preview leads 71.7 to 52.2.
  • The biggest single-benchmark swing is SimpleQA Verified: 73.5% for Gemini 3.1 Pro Preview and 30.2% for Grok 4.20 (Non-Reasoning).
  • Grok 4.20 (Non-Reasoning) is cheaper at $1.25 / $2.50 per million input/output tokens, against $2 / $12 for Gemini 3.1 Pro Preview.
  • Gemini 3.1 Pro Preview accepts more context: 1.05M tokens versus 1M.

Side by side

Gemini 3.1 Pro PreviewGrok 4.20 (Non-Reasoning)
ProviderGooglexAI
Noometry Index56.748.5
Released2026-02-192026-02-17
WeightsProprietaryProprietary
Context window1.05M1M
Max output66K30K
Input $ / M tokens$2$1.25
Output $ / M tokens$12$2.50
Results tracked7146

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

Category by category

Coding Too close to call

Gemini 3.1 Pro Preview: 42.7 (#88), Grok 4.20 (Non-Reasoning): 41.8 (#104)

BenchmarkGemini 3.1 Pro PreviewGrok 4.20 (Non-Reasoning)
LMArena WebDev14461374
WeirdML72.1%52.3%
LMArena Coding14831458
ALE-Bench1,1611,150
SWE-bench Verified75.6%—
DeepSWE11.7%—
SciCode58.9%—
GSO22.6%—
MirrorCode8.9%—
AlgoTune2.02—

Agentic & Tool Use Gemini 3.1 Pro Preview leads

Gemini 3.1 Pro Preview: 37.7 (#33), Grok 4.20 (Non-Reasoning): 34.3 (#43)

BenchmarkGemini 3.1 Pro PreviewGrok 4.20 (Non-Reasoning)
Terminal-Bench80.2%57.3%
τ²-bench Banking26%18%
LMArena Search12111189
Vending-Bench 23,7744,663
APEX-Agents35.3%—
DeepResearch Bench47.8%—
PostTrainBench22%—
BALROG57%—
ExploitBench26.1%—
GBAEval0.8%—
GDP.pdf17%—
METR Time Horizons77%—

Reasoning Gemini 3.1 Pro Preview leads

Gemini 3.1 Pro Preview: 71.7 (#9), Grok 4.20 (Non-Reasoning): 52.2 (#30)

BenchmarkGemini 3.1 Pro PreviewGrok 4.20 (Non-Reasoning)
ARC-AGI-277.1%65.1%
NYT Connections (extended)97.4%85.4%
ARC-AGI-198%89.5%
Chess Puzzles55%24%
Thematic Generalization79.4%63.8%
LMArena Hard Prompts14841451
DTBench97.1%90.1%
LMCA53.8%38.7%
Epoch Capabilities Index154.92152.01
ForecastBench5961.4
SimpleBench79.6%—
Kagi LLM Benchmark—75%
CritPt17.7%—
EnigmaEval36.8%—
EBR-Bench14.3%—
Mystery Game Puzzles34%—

Math Gemini 3.1 Pro Preview leads

Gemini 3.1 Pro Preview: 61.8 (#28), Grok 4.20 (Non-Reasoning): 48.0 (#54)

BenchmarkGemini 3.1 Pro PreviewGrok 4.20 (Non-Reasoning)
FrontierMath (Tiers 1-3)59.6%44.9%
FrontierMath Tier 426.8%17.1%
OTIS Mock AIME 2024-202595.6%92.2%
ProofBench26%14%
LMArena Math14851455
MathArena Final-Answer Competitions86.5%—
FrontierMath (Feb 2025 set)36.9%—
FrontierMath Tier 4 (v1)16.7%—

Knowledge Gemini 3.1 Pro Preview leads

Gemini 3.1 Pro Preview: 71.8 (#3), Grok 4.20 (Non-Reasoning): 52.8 (#54)

BenchmarkGemini 3.1 Pro PreviewGrok 4.20 (Non-Reasoning)
GPQA Diamond94.4%89.3%
SimpleQA Verified73.5%30.2%
LMArena Expert14841440
Humanity's Last Exam46.4%—
Vectara Hallucination Rate10.4%—

Multimodal Gemini 3.1 Pro Preview leads

Gemini 3.1 Pro Preview: 38.1 (#60), Grok 4.20 (Non-Reasoning): 33.2 (#88)

BenchmarkGemini 3.1 Pro PreviewGrok 4.20 (Non-Reasoning)
LMArena Vision12961263
Blueprint-Bench 226.5%0%
LMArena Document14441416
Furniture Assembly26.7%—

Multilingual Gemini 3.1 Pro Preview leads

Gemini 3.1 Pro Preview: 57.1 (#11), Grok 4.20 (Non-Reasoning): 54.7 (#38)

BenchmarkGemini 3.1 Pro PreviewGrok 4.20 (Non-Reasoning)
LMArena Non-English14751440
LMArena Chinese15281481
LMArena French14881477
LMArena German14911463
LMArena Japanese14941450
LMArena Korean14581417
LMArena Russian14961457
LMArena Spanish14801442

Instruction Following Gemini 3.1 Pro Preview leads

Gemini 3.1 Pro Preview: 77.1 (#31), Grok 4.20 (Non-Reasoning): 75.0 (#79)

BenchmarkGemini 3.1 Pro PreviewGrok 4.20 (Non-Reasoning)
LMArena Instruction Following14651420

Long Context Gemini 3.1 Pro Preview leads

Gemini 3.1 Pro Preview: 47.3 (#17), Grok 4.20 (Non-Reasoning): 45.4 (#33)

BenchmarkGemini 3.1 Pro PreviewGrok 4.20 (Non-Reasoning)
CL-bench20.8%22.2%
CL-bench Life16.9%11.9%
LMArena Longer Query14821436

Writing & Preference Too close to call

Gemini 3.1 Pro Preview: 66.3 (#36), Grok 4.20 (Non-Reasoning): 65.9 (#42)

BenchmarkGemini 3.1 Pro PreviewGrok 4.20 (Non-Reasoning)
LMArena Text14801451
LMArena Creative Writing14821439
EQ-Bench Creative Writing14911574
LMArena Multi-Turn14861456
EQ-Bench 41142—

Frequently asked questions

Is Gemini 3.1 Pro Preview better than Grok 4.20 (Non-Reasoning)?

Gemini 3.1 Pro Preview is the stronger model overall, scoring 56.7 to 48.5 on the Noometry Index. Grok 4.20 (Non-Reasoning) costs 2.9× less per token, which makes it the better buy when Gemini 3.1 Pro Preview's lead doesn't matter for your workload.

Which is cheaper, Gemini 3.1 Pro Preview or Grok 4.20 (Non-Reasoning)?

Grok 4.20 (Non-Reasoning) is cheaper. It lists at $1.25 per million input tokens and $2.50 per million output tokens; Gemini 3.1 Pro Preview lists at $2 and $12.

Is Gemini 3.1 Pro Preview or Grok 4.20 (Non-Reasoning) better for coding?

They score almost the same on coding (42.7 vs 41.8); test both on your own repository before choosing.

Which has the bigger context window?

Gemini 3.1 Pro Preview does, with 1.05M tokens against 1M.

How many benchmarks do Gemini 3.1 Pro Preview and Grok 4.20 (Non-Reasoning) share?

45 benchmarks have published results for both models. Gemini 3.1 Pro Preview has 71 scored results on Noometry and Grok 4.20 (Non-Reasoning) has 46.

Related comparisons

Go deeper