Noometry

Model comparison

Claude Opus 4.8 vs Gemini 3.1 Pro Preview

Claude Opus 4.8 is the stronger model overall, scoring 60.7 to 56.7 on the Noometry Index. Gemini 3.1 Pro Preview costs 2.2× less per token, which makes it the better buy when Claude Opus 4.8's lead doesn't matter for your workload.

Last verified . 59 shared benchmarks.

Claude Opus 4.8 Anthropic

60.7

Rank #9 Confirmed

Gemini 3.1 Pro Preview Google

56.7

Rank #20 Confirmed

Summary

  • They share 59 benchmarks with published results for both. Claude Opus 4.8 scores higher in 6 categories and Gemini 3.1 Pro Preview in 4 categories; 9 gaps are clear of the uncertainty.
  • The widest gap is in coding, where Claude Opus 4.8 leads 60.2 to 42.7.
  • The biggest single-benchmark swing is GBAEval: 70.9% for Claude Opus 4.8 and 0.8% for Gemini 3.1 Pro Preview.
  • Gemini 3.1 Pro Preview is cheaper at $2 / $12 per million input/output tokens, against $5 / $25 for Claude Opus 4.8.
  • Gemini 3.1 Pro Preview accepts more context: 1.05M tokens versus 1M.

Side by side

Claude Opus 4.8Gemini 3.1 Pro Preview
ProviderAnthropicGoogle
Noometry Index60.756.7
Released2026-05-282026-02-19
WeightsProprietaryProprietary
Context window1M1.05M
Max output128K66K
Input $ / M tokens$5$2
Output $ / M tokens$25$12
Results tracked6571

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

Category by category

Coding Claude Opus 4.8 leads

Claude Opus 4.8: 60.2 (#9), Gemini 3.1 Pro Preview: 42.7 (#88)

BenchmarkClaude Opus 4.8Gemini 3.1 Pro Preview
DeepSWE59%11.7%
LMArena WebDev15561446
SciCode53.5%58.9%
GSO47.1%22.6%
WeirdML82.9%72.1%
LMArena Coding14911483
ALE-Bench1,5641,161
SWE-bench Verified—75.6%
FrontierCode46.5%—
MirrorCode—8.9%
AlgoTune—2.02

Agentic & Tool Use Claude Opus 4.8 leads

Claude Opus 4.8: 47.6 (#10), Gemini 3.1 Pro Preview: 37.7 (#33)

BenchmarkClaude Opus 4.8Gemini 3.1 Pro Preview
APEX-Agents48.9%35.3%
τ²-bench Banking39.7%26%
DeepResearch Bench50.2%47.8%
PostTrainBench33.8%22%
GBAEval70.9%0.8%
GDP.pdf24%17%
LMArena Search12041211
Vending-Bench 25,7873,774
Terminal-Bench—80.2%
OSWorld 2.020.6%—
Remote Labor Index8.3%—
BALROG—57%
ExploitBench—26.1%
METR Time Horizons—77%

Reasoning Gemini 3.1 Pro Preview leads

Claude Opus 4.8: 64.6 (#14), Gemini 3.1 Pro Preview: 71.7 (#9)

BenchmarkClaude Opus 4.8Gemini 3.1 Pro Preview
ARC-AGI-272.1%77.1%
SimpleBench64.8%79.6%
NYT Connections (extended)91.1%97.4%
ARC-AGI-192.5%98%
CritPt20.9%17.7%
Chess Puzzles34%55%
EnigmaEval23.5%36.8%
EBR-Bench28.6%14.3%
LMArena Hard Prompts14811484
Mystery Game Puzzles36%34%
DTBench94.9%97.1%
LMCA57.5%53.8%
Epoch Capabilities Index158.3154.92
ForecastBench59.959
Kagi LLM Benchmark88.8%—
Thematic Generalization—79.4%
Surface Evolver Bench87.5%—
Bench to the Future 30.14—

Math Claude Opus 4.8 leads

Claude Opus 4.8: 78.2 (#11), Gemini 3.1 Pro Preview: 61.8 (#28)

BenchmarkClaude Opus 4.8Gemini 3.1 Pro Preview
FrontierMath (Tiers 1-3)80%59.6%
FrontierMath Tier 456.1%26.8%
MathArena Final-Answer Competitions91.8%86.5%
OTIS Mock AIME 2024-202598.3%95.6%
ProofBench69%26%
LMArena Math14871485
FrontierMath (Feb 2025 set)47.2%36.9%
FrontierMath Tier 4 (v1)31.3%16.7%

Knowledge Gemini 3.1 Pro Preview leads

Claude Opus 4.8: 61.3 (#26), Gemini 3.1 Pro Preview: 71.8 (#3)

BenchmarkClaude Opus 4.8Gemini 3.1 Pro Preview
GPQA Diamond91%94.4%
SimpleQA Verified53%73.5%
LMArena Expert15011484
Humanity's Last Exam—46.4%
Vectara Hallucination Rate—10.4%

Multimodal Claude Opus 4.8 leads

Claude Opus 4.8: 43.2 (#20), Gemini 3.1 Pro Preview: 38.1 (#60)

BenchmarkClaude Opus 4.8Gemini 3.1 Pro Preview
LMArena Vision12951296
Blueprint-Bench 214.5%26.5%
Furniture Assembly42.5%26.7%
LMArena Document14751444

Multilingual Gemini 3.1 Pro Preview leads

Claude Opus 4.8: 55.4 (#31), Gemini 3.1 Pro Preview: 57.1 (#11)

BenchmarkClaude Opus 4.8Gemini 3.1 Pro Preview
LMArena Non-English14501475
LMArena Chinese15091528
LMArena French14821488
LMArena German14701491
LMArena Japanese14401494
LMArena Korean14321458
LMArena Russian14761496
LMArena Spanish14701480

Instruction Following Too close to call

Claude Opus 4.8: 77.6 (#22), Gemini 3.1 Pro Preview: 77.1 (#31)

BenchmarkClaude Opus 4.8Gemini 3.1 Pro Preview
LMArena Instruction Following14761465

Long Context Gemini 3.1 Pro Preview leads

Claude Opus 4.8: 45.4 (#37), Gemini 3.1 Pro Preview: 47.3 (#17)

BenchmarkClaude Opus 4.8Gemini 3.1 Pro Preview
LMArena Longer Query14821482
CL-bench—20.8%
CL-bench Life—16.9%

Writing & Preference Claude Opus 4.8 leads

Claude Opus 4.8: 72.1 (#17), Gemini 3.1 Pro Preview: 66.3 (#36)

BenchmarkClaude Opus 4.8Gemini 3.1 Pro Preview
LMArena Text14601480
LMArena Creative Writing14531482
EQ-Bench Creative Writing18401491
EQ-Bench 412811142
LMArena Multi-Turn14761486

Frequently asked questions

Is Claude Opus 4.8 better than Gemini 3.1 Pro Preview?

Claude Opus 4.8 is the stronger model overall, scoring 60.7 to 56.7 on the Noometry Index. Gemini 3.1 Pro Preview costs 2.2× less per token, which makes it the better buy when Claude Opus 4.8's lead doesn't matter for your workload.

Which is cheaper, Claude Opus 4.8 or Gemini 3.1 Pro Preview?

Gemini 3.1 Pro Preview is cheaper. It lists at $2 per million input tokens and $12 per million output tokens; Claude Opus 4.8 lists at $5 and $25.

Is Claude Opus 4.8 or Gemini 3.1 Pro Preview better for coding?

Claude Opus 4.8 scores higher on coding benchmarks: 60.2 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 1M.

How many benchmarks do Claude Opus 4.8 and Gemini 3.1 Pro Preview share?

59 benchmarks have published results for both models. Claude Opus 4.8 has 65 scored results on Noometry and Gemini 3.1 Pro Preview has 71.

Related comparisons

Go deeper