Model comparison
Claude Opus 4.6 vs Gemini 3 Pro
Claude Opus 4.6 is the stronger model overall, scoring 58.3 to 54.7 on the Noometry Index.
Last verified . 53 shared benchmarks.
GoogleSummary
- They share 53 benchmarks with published results for both. Claude Opus 4.6 scores higher in 8 categories and Gemini 3 Pro in 2 categories; 10 gaps are clear of the uncertainty.
- The widest gap is in multimodal, where Gemini 3 Pro leads 56.9 to 37.0.
- The biggest single-benchmark swing is ARC-AGI-2: 69.2% for Claude Opus 4.6 and 31.1% for Gemini 3 Pro.
Side by side
| Claude Opus 4.6 | Gemini 3 Pro | |
|---|---|---|
| Provider | Anthropic | |
| Noometry Index | 58.3 | 54.7 |
| Released | 2026-02-04 | 2025-11-18 |
| Weights | Proprietary | Proprietary |
| Context window | 1M | — |
| Max output | 128K | — |
| Input $ / M tokens | $5 | — |
| Output $ / M tokens | $25 | — |
| Results tracked | 68 | 67 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Claude Opus 4.6 leads
Claude Opus 4.6: 57.7 (#15), Gemini 3 Pro: 52.1 (#33)
| Benchmark | Claude Opus 4.6 | Gemini 3 Pro |
|---|---|---|
| SWE-bench Verified | 78.7% | 72.9% |
| SWE-bench Verified (bash only) | 75.6% | 74.2% |
| LMArena WebDev | 1546 | 1439 |
| SWE-bench Multilingual | 72% | 68.7% |
| GSO | 41.2% | 18.6% |
| WeirdML | 78% | 69.9% |
| LMArena Coding | 1535 | 1482 |
| ALE-Bench | 996.5 | 1,177 |
| AlgoTune | 1.47 | 1.83 |
| FrontierCode | 26.6% | — |
Agentic & Tool Use Claude Opus 4.6 leads
Claude Opus 4.6: 51.0 (#4), Gemini 3 Pro: 40.5 (#21)
| Benchmark | Claude Opus 4.6 | Gemini 3 Pro |
|---|---|---|
| Terminal-Bench | 79.8% | 69.4% |
| Remote Labor Index | 4.2% | 1.3% |
| τ²-bench Banking | 27.3% | 18% |
| DeepResearch Bench | 55.3% | 46.3% |
| LMArena Search | 1253 | 1207 |
| METR Time Horizons | 78.9% | 71% |
| Vending-Bench 2 | 8,018 | 5,478 |
| APEX-Agents | 46.3% | — |
| Berkeley Function Calling Leaderboard | — | 72.5% |
| GDPval | — | 40.3% |
| τ²-bench Airline | — | 80.5% |
| τ²-bench Retail | — | 75.9% |
| τ²-bench Telecom | — | 91% |
| Cybench | 93% | — |
| BALROG | — | 58.1% |
| GBAEval | 44.1% | — |
Reasoning Claude Opus 4.6 leads
Claude Opus 4.6: 57.7 (#21), Gemini 3 Pro: 52.5 (#28)
| Benchmark | Claude Opus 4.6 | Gemini 3 Pro |
|---|---|---|
| ARC-AGI-2 | 69.2% | 31.1% |
| SimpleBench | 67.6% | 76.4% |
| Kagi LLM Benchmark | 83.6% | 80.1% |
| NYT Connections (extended) | 92.1% | 94.4% |
| ARC-AGI-1 | 94% | 75% |
| Chess Puzzles | 17% | 31% |
| EnigmaEval | 7.6% | 18.2% |
| LMArena Hard Prompts | 1527 | 1481 |
| Epoch Capabilities Index | 155.36 | 152.96 |
| ForecastBench | 60 | 61.2 |
| CritPt | — | 6.9% |
| Thematic Generalization | 80.6% | — |
| EBR-Bench | 12.7% | — |
| Mystery Game Puzzles | 25% | — |
| DTBench | 91.2% | — |
| LMCA | 55.8% | — |
Math Claude Opus 4.6 leads
Claude Opus 4.6: 62.7 (#25), Gemini 3 Pro: 48.8 (#50)
| Benchmark | Claude Opus 4.6 | Gemini 3 Pro |
|---|---|---|
| MathArena Final-Answer Competitions | 78.5% | 67% |
| OTIS Mock AIME 2024-2025 | 94.4% | 91.4% |
| ProofBench | 50% | 20% |
| LMArena Math | 1517 | 1475 |
| FrontierMath (Feb 2025 set) | 40.7% | 37.6% |
| FrontierMath Tier 4 (v1) | 22.9% | 18.8% |
| FrontierMath (Tiers 1-3) | 66% | — |
| FrontierMath Tier 4 | 26.8% | — |
| Omni-MATH | — | 55.5% |
Knowledge Gemini 3 Pro leads
Claude Opus 4.6: 61.8 (#23), Gemini 3 Pro: 64.3 (#15)
| Benchmark | Claude Opus 4.6 | Gemini 3 Pro |
|---|---|---|
| GPQA Diamond | 90.5% | 92.6% |
| Humanity's Last Exam | 34.4% | 37.5% |
| Vectara Hallucination Rate | 12.2% | 13.6% |
| LMArena Expert | 1546 | 1475 |
| SimpleQA Verified | 47% | — |
| MMLU-Pro | — | 90.3% |
| GPQA (HELM) | — | 80.3% |
Multimodal Gemini 3 Pro leads
Claude Opus 4.6: 37.0 (#67), Gemini 3 Pro: 56.9 (#1)
| Benchmark | Claude Opus 4.6 | Gemini 3 Pro |
|---|---|---|
| LMArena Vision | 1316 | 1305 |
| LMArena Document | 1507 | 1434 |
| GeoBench | — | 84% |
| VPCT | — | 91% |
| Furniture Assembly | 28.3% | — |
Multilingual Claude Opus 4.6 leads
Claude Opus 4.6: 58.2 (#4), Gemini 3 Pro: 57.1 (#14)
| Benchmark | Claude Opus 4.6 | Gemini 3 Pro |
|---|---|---|
| LMArena Non-English | 1490 | 1474 |
| LMArena Chinese | 1551 | 1522 |
| LMArena French | 1513 | 1492 |
| LMArena German | 1504 | 1517 |
| LMArena Japanese | 1484 | 1513 |
| LMArena Korean | 1465 | 1449 |
| LMArena Russian | 1497 | 1493 |
| LMArena Spanish | 1512 | 1468 |
Instruction Following Claude Opus 4.6 leads
Claude Opus 4.6: 79.7 (#3), Gemini 3 Pro: 76.4 (#42)
| Benchmark | Claude Opus 4.6 | Gemini 3 Pro |
|---|---|---|
| LMArena Instruction Following | 1523 | 1458 |
| IFEval | — | 87.7% |
Long Context Claude Opus 4.6 leads
Claude Opus 4.6: 48.1 (#13), Gemini 3 Pro: 44.0 (#75)
| Benchmark | Claude Opus 4.6 | Gemini 3 Pro |
|---|---|---|
| CL-bench | 20.7% | 15.8% |
| LMArena Longer Query | 1520 | 1472 |
| CL-bench Life | 17% | — |
Writing & Preference Claude Opus 4.6 leads
Claude Opus 4.6: 73.7 (#9), Gemini 3 Pro: 66.6 (#34)
| Benchmark | Claude Opus 4.6 | Gemini 3 Pro |
|---|---|---|
| LMArena Text | 1504 | 1479 |
| LMArena Creative Writing | 1506 | 1482 |
| EQ-Bench Creative Writing | 1809 | 1525 |
| LMArena Multi-Turn | 1514 | 1484 |
| WildBench | — | 85.9% |
| EQ-Bench 4 | 1223 | — |
Frequently asked questions
Is Claude Opus 4.6 better than Gemini 3 Pro?
Claude Opus 4.6 is the stronger model overall, scoring 58.3 to 54.7 on the Noometry Index.
Is Claude Opus 4.6 or Gemini 3 Pro better for coding?
Claude Opus 4.6 scores higher on coding benchmarks: 57.7 versus 52.1 in the Noometry coding category.
How many benchmarks do Claude Opus 4.6 and Gemini 3 Pro share?
53 benchmarks have published results for both models. Claude Opus 4.6 has 68 scored results on Noometry and Gemini 3 Pro has 67.