Model comparison
Gemini 3 Pro vs Grok 4.5
Gemini 3 Pro and Grok 4.5 score almost the same on the Noometry Index (54.8 vs 55.0), so choose on price, context window or the category you care about most.
Last verified . 37 shared benchmarks.
GoogleSummary
- They share 37 benchmarks with published results for both. Gemini 3 Pro scores higher in 5 categories and Grok 4.5 in 5 categories; 6 gaps are clear of the uncertainty.
- The widest gap is in multimodal, where Gemini 3 Pro leads 57.6 to 37.6.
- The biggest single-benchmark swing is τ²-bench Banking: 18% for Gemini 3 Pro and 47.9% for Grok 4.5.
Side by side
| Gemini 3 Pro | Grok 4.5 | |
|---|---|---|
| Provider | xAI | |
| Noometry Index | 54.8 | 55.0 |
| Released | 2025-11-18 | 2026-07-08 |
| Weights | Proprietary | Proprietary |
| Context window | — | 500K |
| Max output | — | 500K |
| Input $ / M tokens | — | $2 |
| Output $ / M tokens | — | $6 |
| Results tracked | 67 | 52 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Too close to call
Gemini 3 Pro: 51.6 (#39), Grok 4.5: 52.2 (#35)
| Benchmark | Gemini 3 Pro | Grok 4.5 |
|---|---|---|
| LMArena WebDev | 1440 | 1553 |
| WeirdML | 69.9% | 46.4% |
| LMArena Coding | 1481 | 1474 |
| ALE-Bench | 1,177 | 1,309 |
| SWE-bench Verified | 72.9% | — |
| DeepSWE | — | 53.8% |
| FrontierCode | — | 42.4% |
| SWE-bench Verified (bash only) | 74.2% | — |
| SWE-bench Multilingual | 68.7% | — |
| SciCode | — | 54.1% |
| GSO | 18.6% | — |
| AlgoTune | 1.83 | — |
Agentic & Tool Use Grok 4.5 leads
Gemini 3 Pro: 40.6 (#23), Grok 4.5: 44.4 (#17)
| Benchmark | Gemini 3 Pro | Grok 4.5 |
|---|---|---|
| τ²-bench Banking | 18% | 47.9% |
| LMArena Search | 1207 | 1213 |
| Vending-Bench 2 | 5,478 | 3,887 |
| Terminal-Bench | 69.4% | — |
| APEX-Agents | — | 56.2% |
| Berkeley Function Calling Leaderboard | 72.5% | — |
| GDPval | 40.3% | — |
| Remote Labor Index | 1.3% | — |
| τ²-bench Airline | 80.5% | — |
| τ²-bench Retail | 75.9% | — |
| τ²-bench Telecom | 91% | — |
| DeepResearch Bench | 46.3% | — |
| PostTrainBench | — | 23.4% |
| BALROG | 58.1% | — |
| GBAEval | — | 65.4% |
| GDP.pdf | — | 14% |
| METR Time Horizons | 71% | — |
Reasoning Grok 4.5 leads
Gemini 3 Pro: 52.5 (#31), Grok 4.5: 56.1 (#25)
| Benchmark | Gemini 3 Pro | Grok 4.5 |
|---|---|---|
| ARC-AGI-2 | 31.1% | 52.6% |
| SimpleBench | 76.4% | 70% |
| Kagi LLM Benchmark | 80.1% | 83.5% |
| NYT Connections (extended) | 94.4% | 79.9% |
| ARC-AGI-1 | 75% | 87.2% |
| CritPt | 6.9% | 15.4% |
| Chess Puzzles | 31% | 36% |
| LMArena Hard Prompts | 1480 | 1462 |
| Epoch Capabilities Index | 152.92 | 153.92 |
| EnigmaEval | 18.2% | — |
| DTBench | — | 96.5% |
| LMCA | — | 45.2% |
| Surface Evolver Bench | — | 74.4% |
| ForecastBench | 61.2 | — |
Math Grok 4.5 leads
Gemini 3 Pro: 49.9 (#59), Grok 4.5: 60.9 (#35)
| Benchmark | Gemini 3 Pro | Grok 4.5 |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 91.4% | 97.8% |
| ProofBench | 20% | 31% |
| LMArena Math | 1476 | 1459 |
| FrontierMath (Tiers 1-3) | — | 57.2% |
| FrontierMath Tier 4 | — | 24.4% |
| MathArena Final-Answer Competitions | 67% | — |
| Omni-MATH | 55.5% | — |
| FrontierMath (Feb 2025 set) | 37.6% | — |
| FrontierMath Tier 4 (v1) | 18.8% | — |
Knowledge Gemini 3 Pro leads
Gemini 3 Pro: 64.4 (#16), Grok 4.5: 62.3 (#24)
| Benchmark | Gemini 3 Pro | Grok 4.5 |
|---|---|---|
| GPQA Diamond | 92.6% | 93.4% |
| LMArena Expert | 1475 | 1466 |
| Humanity's Last Exam | 37.5% | — |
| SimpleQA Verified | — | 48.3% |
| MMLU-Pro | 90.3% | — |
| Vectara Hallucination Rate | 13.6% | — |
| GPQA (HELM) | 80.3% | — |
Multimodal Gemini 3 Pro leads
Gemini 3 Pro: 57.6 (#2), Grok 4.5: 37.6 (#72)
| Benchmark | Gemini 3 Pro | Grok 4.5 |
|---|---|---|
| LMArena Vision | 1305 | 1288 |
| LMArena Document | 1434 | 1452 |
| GeoBench | 84% | — |
| VPCT | 91% | — |
| Blueprint-Bench 2 | — | 27.3% |
| Furniture Assembly | — | 22.5% |
Multilingual Gemini 3 Pro leads
Gemini 3 Pro: 56.9 (#16), Grok 4.5: 54.4 (#42)
| Benchmark | Gemini 3 Pro | Grok 4.5 |
|---|---|---|
| LMArena Non-English | 1474 | 1440 |
| LMArena Chinese | 1523 | 1496 |
| LMArena French | 1492 | 1456 |
| LMArena German | 1515 | 1446 |
| LMArena Japanese | 1510 | 1428 |
| LMArena Korean | 1448 | 1404 |
| LMArena Russian | 1493 | 1448 |
| LMArena Spanish | 1470 | 1450 |
Instruction Following Too close to call
Gemini 3 Pro: 76.3 (#45), Grok 4.5: 76.0 (#48)
| Benchmark | Gemini 3 Pro | Grok 4.5 |
|---|---|---|
| LMArena Instruction Following | 1458 | 1446 |
| IFEval | 87.7% | — |
Long Context Too close to call
Gemini 3 Pro: 44.0 (#79), Grok 4.5: 44.8 (#56)
| Benchmark | Gemini 3 Pro | Grok 4.5 |
|---|---|---|
| LMArena Longer Query | 1471 | 1463 |
| CL-bench | 15.8% | — |
Writing & Preference Too close to call
Gemini 3 Pro: 66.4 (#35), Grok 4.5: 65.8 (#42)
| Benchmark | Gemini 3 Pro | Grok 4.5 |
|---|---|---|
| LMArena Text | 1479 | 1448 |
| LMArena Creative Writing | 1482 | 1442 |
| EQ-Bench Creative Writing | 1525 | 1579 |
| LMArena Multi-Turn | 1484 | 1456 |
| WildBench | 85.9% | — |
Frequently asked questions
Is Gemini 3 Pro better than Grok 4.5?
Gemini 3 Pro and Grok 4.5 score almost the same on the Noometry Index (54.8 vs 55.0), so choose on price, context window or the category you care about most.
Is Gemini 3 Pro or Grok 4.5 better for coding?
They score almost the same on coding (51.6 vs 52.2); test both on your own repository before choosing.
How many benchmarks do Gemini 3 Pro and Grok 4.5 share?
37 benchmarks have published results for both models. Gemini 3 Pro has 67 scored results on Noometry and Grok 4.5 has 52.