Noometry

Model comparison

DeepSeek-R1 vs Gemini 3.1 Pro Preview

Gemini 3.1 Pro Preview is the stronger model overall, scoring 56.7 to 42.0 on the Noometry Index. DeepSeek-R1 costs 4.9× 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 . 34 shared benchmarks.

DeepSeek-R1 DeepSeek

42.0

Rank #110 Confirmed

Gemini 3.1 Pro Preview Google

56.7

Rank #20 Confirmed

Summary

  • They share 34 benchmarks with published results for both. DeepSeek-R1 scores higher in 1 category and Gemini 3.1 Pro Preview in 8 categories; 9 gaps are clear of the uncertainty.
  • The widest gap is in reasoning, where Gemini 3.1 Pro Preview leads 71.7 to 18.7.
  • The biggest single-benchmark swing is ARC-AGI-1: 21.2% for DeepSeek-R1 and 98% for Gemini 3.1 Pro Preview.
  • DeepSeek-R1 is cheaper at $0.50 / $2.15 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 164K.

Side by side

DeepSeek-R1Gemini 3.1 Pro Preview
ProviderDeepSeekGoogle
Noometry Index42.056.7
Released2025-01-202026-02-19
WeightsProprietaryProprietary
Context window164K1.05M
Max output64K66K
Input $ / M tokens$0.50$2
Output $ / M tokens$2.15$12
Results tracked5271

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

Category by category

Coding DeepSeek-R1 leads

DeepSeek-R1: 45.9 (#63), Gemini 3.1 Pro Preview: 42.7 (#88)

BenchmarkDeepSeek-R1Gemini 3.1 Pro Preview
SciCode35.7%58.9%
WeirdML41.6%72.1%
LMArena Coding14271483
ALE-Bench804.121,161
AlgoTune1.72.02
SWE-bench Verified—75.6%
DeepSWE—11.7%
Aider Polyglot71.4%—
LMArena WebDev—1446
GSO—22.6%
LiveBench Coding66.7%—
MirrorCode—8.9%

Agentic & Tool Use Gemini 3.1 Pro Preview leads

DeepSeek-R1: 30.4 (#68), Gemini 3.1 Pro Preview: 37.7 (#33)

BenchmarkDeepSeek-R1Gemini 3.1 Pro Preview
DeepResearch Bench35.1%47.8%
BALROG34.9%57%
METR Time Horizons53.8%77%
Terminal-Bench—80.2%
APEX-Agents—35.3%
τ²-bench Banking—26%
PostTrainBench—22%
ExploitBench—26.1%
GBAEval—0.8%
GDP.pdf—17%
LMArena Search—1211
Vending-Bench 2—3,774

Reasoning Gemini 3.1 Pro Preview leads

DeepSeek-R1: 18.7 (#253), Gemini 3.1 Pro Preview: 71.7 (#9)

BenchmarkDeepSeek-R1Gemini 3.1 Pro Preview
ARC-AGI-21.3%77.1%
SimpleBench40.8%79.6%
ARC-AGI-121.2%98%
CritPt1.1%17.7%
LMArena Hard Prompts14151484
Epoch Capabilities Index141.3154.92
ForecastBench6059
Kagi LLM Benchmark69.4%—
NYT Connections (extended)—97.4%
Chess Puzzles—55%
EnigmaEval—36.8%
Thematic Generalization—79.4%
EBR-Bench—14.3%
LiveBench Reasoning83.2%—
Mystery Game Puzzles—34%
DTBench—97.1%
LiveBench Data Analysis69.8%—
LMCA—53.8%
LiveBench71.6%—

Math Gemini 3.1 Pro Preview leads

DeepSeek-R1: 42.0 (#70), Gemini 3.1 Pro Preview: 61.8 (#28)

BenchmarkDeepSeek-R1Gemini 3.1 Pro Preview
OTIS Mock AIME 2024-202566.4%95.6%
LMArena Math14011485
FrontierMath (Tiers 1-3)—59.6%
FrontierMath Tier 4—26.8%
MathArena Final-Answer Competitions—86.5%
ProofBench—26%
Omni-MATH42.4%—
LiveBench Math80.7%—
MATH Level 596.6%—
FrontierMath (Feb 2025 set)—36.9%
FrontierMath Tier 4 (v1)—16.7%

Knowledge Gemini 3.1 Pro Preview leads

DeepSeek-R1: 44.3 (#81), Gemini 3.1 Pro Preview: 71.8 (#3)

BenchmarkDeepSeek-R1Gemini 3.1 Pro Preview
GPQA Diamond76.3%94.4%
Vectara Hallucination Rate11.3%10.4%
LMArena Expert13941484
Humanity's Last Exam—46.4%
SimpleQA Verified—73.5%
MMLU-Pro79.3%—
Confabulations12.7%—
GPQA (HELM)66.6%—

Multimodal Not comparable

DeepSeek-R1: —, Gemini 3.1 Pro Preview: 38.1 (#60)

BenchmarkDeepSeek-R1Gemini 3.1 Pro Preview
LMArena Vision—1296
Blueprint-Bench 2—26.5%
Furniture Assembly—26.7%
LMArena Document—1444

Multilingual Gemini 3.1 Pro Preview leads

DeepSeek-R1: 52.6 (#80), Gemini 3.1 Pro Preview: 57.1 (#11)

BenchmarkDeepSeek-R1Gemini 3.1 Pro Preview
LMArena Non-English14121475
LMArena Chinese14431528
LMArena French14151488
LMArena German14051491
LMArena Japanese13941494
LMArena Korean13591458
LMArena Russian14231496
LMArena Spanish14101480

Instruction Following Gemini 3.1 Pro Preview leads

DeepSeek-R1: 72.2 (#138), Gemini 3.1 Pro Preview: 77.1 (#31)

BenchmarkDeepSeek-R1Gemini 3.1 Pro Preview
LMArena Instruction Following13821465
LiveBench Instruction Following80.5%—
IFEval78.4%—

Long Context Gemini 3.1 Pro Preview leads

DeepSeek-R1: 45.4 (#36), Gemini 3.1 Pro Preview: 47.3 (#17)

BenchmarkDeepSeek-R1Gemini 3.1 Pro Preview
LMArena Longer Query13911482
Fiction.LiveBench75%—
CL-bench—20.8%
CL-bench Life—16.9%

Writing & Preference Gemini 3.1 Pro Preview leads

DeepSeek-R1: 61.5 (#84), Gemini 3.1 Pro Preview: 66.3 (#36)

BenchmarkDeepSeek-R1Gemini 3.1 Pro Preview
LMArena Text14281480
LMArena Creative Writing14061482
EQ-Bench Creative Writing15001491
LMArena Multi-Turn14041486
Short-Story Creative Writing83%—
WildBench82.8%—
EQ-Bench 4—1142
LiveBench Language48.5%—

Frequently asked questions

Is DeepSeek-R1 better than Gemini 3.1 Pro Preview?

Gemini 3.1 Pro Preview is the stronger model overall, scoring 56.7 to 42.0 on the Noometry Index. DeepSeek-R1 costs 4.9× 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, DeepSeek-R1 or Gemini 3.1 Pro Preview?

DeepSeek-R1 is cheaper. It lists at $0.50 per million input tokens and $2.15 per million output tokens; Gemini 3.1 Pro Preview lists at $2 and $12.

Is DeepSeek-R1 or Gemini 3.1 Pro Preview better for coding?

DeepSeek-R1 scores higher on coding benchmarks: 45.9 versus 42.7 in the Noometry coding category.

Which has the bigger context window?

Gemini 3.1 Pro Preview does, with 1.05M tokens against 164K.

How many benchmarks do DeepSeek-R1 and Gemini 3.1 Pro Preview share?

34 benchmarks have published results for both models. DeepSeek-R1 has 52 scored results on Noometry and Gemini 3.1 Pro Preview has 71.

Related comparisons

Go deeper