Model comparison
Claude Sonnet 4.6 vs Gemini 3 Pro
Gemini 3 Pro is the stronger model overall, scoring 54.7 to 50.1 on the Noometry Index.
Last verified . 41 shared benchmarks.
GoogleSummary
- They share 41 benchmarks with published results for both. Claude Sonnet 4.6 scores higher in 4 categories and Gemini 3 Pro in 6 categories; 10 gaps are clear of the uncertainty.
- The widest gap is in multimodal, where Gemini 3 Pro leads 56.9 to 38.0.
- The biggest single-benchmark swing is ARC-AGI-2: 60.4% for Claude Sonnet 4.6 and 31.1% for Gemini 3 Pro.
Side by side
| Claude Sonnet 4.6 | Gemini 3 Pro | |
|---|---|---|
| Provider | Anthropic | |
| Noometry Index | 50.1 | 54.7 |
| Released | 2026-02-17 | 2025-11-18 |
| Weights | Proprietary | Proprietary |
| Context window | 1M | — |
| Max output | 128K | — |
| Input $ / M tokens | $3 | — |
| Output $ / M tokens | $15 | — |
| Results tracked | 57 | 67 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Gemini 3 Pro leads
Claude Sonnet 4.6: 46.0 (#62), Gemini 3 Pro: 52.1 (#33)
| Benchmark | Claude Sonnet 4.6 | Gemini 3 Pro |
|---|---|---|
| SWE-bench Verified | 75.2% | 72.9% |
| LMArena WebDev | 1521 | 1439 |
| WeirdML | 66.1% | 69.9% |
| LMArena Coding | 1504 | 1482 |
| ALE-Bench | 1,327 | 1,177 |
| DeepSWE | 29.9% | — |
| FrontierCode | 24.3% | — |
| SWE-bench Verified (bash only) | — | 74.2% |
| SWE-bench Multilingual | — | 68.7% |
| SciCode | 46.8% | — |
| GSO | — | 18.6% |
| AlgoTune | — | 1.83 |
Agentic & Tool Use Gemini 3 Pro leads
Claude Sonnet 4.6: 39.1 (#26), Gemini 3 Pro: 40.5 (#21)
| Benchmark | Claude Sonnet 4.6 | Gemini 3 Pro |
|---|---|---|
| Terminal-Bench | 53.4% | 69.4% |
| DeepResearch Bench | 54.9% | 46.3% |
| LMArena Search | 1221 | 1207 |
| Vending-Bench 2 | 7,204 | 5,478 |
| APEX-Agents | 43% | — |
| Berkeley Function Calling Leaderboard | — | 72.5% |
| OSWorld 2.0 | 9.3% | — |
| GDPval | — | 40.3% |
| Remote Labor Index | — | 1.3% |
| τ²-bench Airline | — | 80.5% |
| τ²-bench Banking | — | 18% |
| τ²-bench Retail | — | 75.9% |
| τ²-bench Telecom | — | 91% |
| OSWorld | 72.1% | — |
| BALROG | — | 58.1% |
| ExploitBench | 23.6% | — |
| GBAEval | 48.8% | — |
| GDP.pdf | 18% | — |
| METR Time Horizons | — | 71% |
Reasoning Gemini 3 Pro leads
Claude Sonnet 4.6: 46.1 (#40), Gemini 3 Pro: 52.5 (#28)
| Benchmark | Claude Sonnet 4.6 | Gemini 3 Pro |
|---|---|---|
| ARC-AGI-2 | 60.4% | 31.1% |
| NYT Connections (extended) | 80.9% | 94.4% |
| ARC-AGI-1 | 86.5% | 75% |
| CritPt | 3.1% | 6.9% |
| Chess Puzzles | 13% | 31% |
| LMArena Hard Prompts | 1485 | 1481 |
| Epoch Capabilities Index | 152.25 | 152.96 |
| ForecastBench | 62 | 61.2 |
| SimpleBench | — | 76.4% |
| Kagi LLM Benchmark | — | 80.1% |
| EnigmaEval | — | 18.2% |
| Thematic Generalization | 76.3% | — |
| Mystery Game Puzzles | 16% | — |
| DTBench | 89.9% | — |
| LMCA | 46.5% | — |
Math Claude Sonnet 4.6 leads
Claude Sonnet 4.6: 51.9 (#41), Gemini 3 Pro: 48.8 (#50)
| Benchmark | Claude Sonnet 4.6 | Gemini 3 Pro |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 85.8% | 91.4% |
| ProofBench | 45% | 20% |
| LMArena Math | 1463 | 1475 |
| FrontierMath (Feb 2025 set) | 32.4% | 37.6% |
| FrontierMath Tier 4 (v1) | 8.3% | 18.8% |
| MathArena Final-Answer Competitions | — | 67% |
| Omni-MATH | — | 55.5% |
Knowledge Gemini 3 Pro leads
Claude Sonnet 4.6: 51.6 (#60), Gemini 3 Pro: 64.3 (#15)
| Benchmark | Claude Sonnet 4.6 | Gemini 3 Pro |
|---|---|---|
| GPQA Diamond | 87.4% | 92.6% |
| Vectara Hallucination Rate | 10.6% | 13.6% |
| LMArena Expert | 1500 | 1475 |
| Humanity's Last Exam | — | 37.5% |
| SimpleQA Verified | 35.5% | — |
| MMLU-Pro | — | 90.3% |
| GPQA (HELM) | — | 80.3% |
Multimodal Gemini 3 Pro leads
Claude Sonnet 4.6: 38.0 (#61), Gemini 3 Pro: 56.9 (#1)
| Benchmark | Claude Sonnet 4.6 | Gemini 3 Pro |
|---|---|---|
| LMArena Vision | 1283 | 1305 |
| LMArena Document | 1482 | 1434 |
| GeoBench | — | 84% |
| VPCT | — | 91% |
| Blueprint-Bench 2 | 6.7% | — |
Multilingual Gemini 3 Pro leads
Claude Sonnet 4.6: 54.6 (#39), Gemini 3 Pro: 57.1 (#14)
| Benchmark | Claude Sonnet 4.6 | Gemini 3 Pro |
|---|---|---|
| LMArena Non-English | 1440 | 1474 |
| LMArena Chinese | 1491 | 1522 |
| LMArena French | 1465 | 1492 |
| LMArena German | 1428 | 1517 |
| LMArena Japanese | 1419 | 1513 |
| LMArena Korean | 1411 | 1449 |
| LMArena Russian | 1440 | 1493 |
| LMArena Spanish | 1465 | 1468 |
Instruction Following Claude Sonnet 4.6 leads
Claude Sonnet 4.6: 77.6 (#23), Gemini 3 Pro: 76.4 (#42)
| Benchmark | Claude Sonnet 4.6 | Gemini 3 Pro |
|---|---|---|
| LMArena Instruction Following | 1476 | 1458 |
| IFEval | — | 87.7% |
Long Context Claude Sonnet 4.6 leads
Claude Sonnet 4.6: 45.3 (#40), Gemini 3 Pro: 44.0 (#75)
| Benchmark | Claude Sonnet 4.6 | Gemini 3 Pro |
|---|---|---|
| LMArena Longer Query | 1479 | 1472 |
| CL-bench | — | 15.8% |
Writing & Preference Claude Sonnet 4.6 leads
Claude Sonnet 4.6: 70.4 (#22), Gemini 3 Pro: 66.6 (#34)
| Benchmark | Claude Sonnet 4.6 | Gemini 3 Pro |
|---|---|---|
| LMArena Text | 1458 | 1479 |
| LMArena Creative Writing | 1435 | 1482 |
| EQ-Bench Creative Writing | 1810 | 1525 |
| LMArena Multi-Turn | 1464 | 1484 |
| WildBench | — | 85.9% |
| EQ-Bench 4 | 1207 | — |
Frequently asked questions
Is Claude Sonnet 4.6 better than Gemini 3 Pro?
Gemini 3 Pro is the stronger model overall, scoring 54.7 to 50.1 on the Noometry Index.
Is Claude Sonnet 4.6 or Gemini 3 Pro better for coding?
Gemini 3 Pro scores higher on coding benchmarks: 52.1 versus 46.0 in the Noometry coding category.
How many benchmarks do Claude Sonnet 4.6 and Gemini 3 Pro share?
41 benchmarks have published results for both models. Claude Sonnet 4.6 has 57 scored results on Noometry and Gemini 3 Pro has 67.