Model comparison
Gemini 2.5 Pro vs Grok 3
Gemini 2.5 Pro is the stronger model overall, scoring 45.0 to 39.9 on the Noometry Index.
Last verified . 40 shared benchmarks.
GoogleSummary
- They share 40 benchmarks with published results for both. Gemini 2.5 Pro scores higher in 6 categories and Grok 3 in 3 categories; 7 gaps are clear of the uncertainty.
- The widest gap is in long context, where Gemini 2.5 Pro leads 59.8 to 38.7.
- The biggest single-benchmark swing is ARC-AGI-1: 41% for Gemini 2.5 Pro and 5.5% for Grok 3.
Side by side
| Gemini 2.5 Pro | Grok 3 | |
|---|---|---|
| Provider | xAI | |
| Noometry Index | 45.0 | 39.9 |
| Released | 2025-03-25 | 2025-04-09 |
| Weights | Proprietary | Proprietary |
| Context window | 1.05M | — |
| Max output | 66K | — |
| Input $ / M tokens | $1.25 | — |
| Output $ / M tokens | $10 | — |
| Results tracked | 78 | 40 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Too close to call
Gemini 2.5 Pro: 42.4 (#101), Grok 3: 41.9 (#115)
| Benchmark | Gemini 2.5 Pro | Grok 3 |
|---|---|---|
| Aider Polyglot | 83.1% | 53.3% |
| WeirdML | 54% | 37.2% |
| LMArena Coding | 1452 | 1432 |
| SWE-bench Verified | 57.6% | — |
| SWE-bench Verified (bash only) | 53.6% | — |
| LMArena WebDev | 1227 | — |
| SciCode | 42.8% | — |
| GSO | 3.9% | — |
| LiveBench Coding | 85.9% | — |
| CadEval | 64% | — |
| ALE-Bench | 785.52 | — |
| AlgoTune | 1.51 | — |
Agentic & Tool Use Grok 3 leads
Gemini 2.5 Pro: 29.2 (#88), Grok 3: 30.5 (#76)
| Benchmark | Gemini 2.5 Pro | Grok 3 |
|---|---|---|
| BALROG | 43.3% | 29.5% |
| Terminal-Bench | 32.6% | — |
| GDPval | 23.3% | — |
| Remote Labor Index | 0.8% | — |
| TheAgentCompany | 30.3% | — |
| τ²-bench Banking | 13.7% | — |
| DeepResearch Bench | 42.8% | — |
| LMArena Search | 1142 | — |
| METR Time Horizons | 55.4% | — |
| Vending-Bench 2 | 573.64 | — |
Reasoning Gemini 2.5 Pro leads
Gemini 2.5 Pro: 28.8 (#99), Grok 3: 13.7 (#333)
| Benchmark | Gemini 2.5 Pro | Grok 3 |
|---|---|---|
| ARC-AGI-2 | 4.9% | 0% |
| SimpleBench | 62.4% | 36.1% |
| Kagi LLM Benchmark | 70.3% | 61.3% |
| ARC-AGI-1 | 41% | 5.5% |
| LMArena Hard Prompts | 1455 | 1434 |
| Epoch Capabilities Index | 145.32 | 138.33 |
| CritPt | 2% | — |
| Chess Puzzles | 20% | — |
| EnigmaEval | 5.6% | — |
| LiveBench Reasoning | 89.8% | — |
| DTBench | 82.4% | — |
| LiveBench Data Analysis | 79.9% | — |
| LMCA | 34.8% | — |
| ForecastBench | 61.3 | — |
| LiveBench | 82.3% | — |
Math Grok 3 leads
Gemini 2.5 Pro: 32.5 (#213), Grok 3: 38.0 (#145)
| Benchmark | Gemini 2.5 Pro | Grok 3 |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 84.7% | 55.6% |
| Omni-MATH | 41.6% | 46.4% |
| LMArena Math | 1450 | 1391 |
| MATH Level 5 | 95.9% | 88.7% |
| FrontierMath (Feb 2025 set) | 14.1% | 3.8% |
| FrontierMath Tier 4 (v1) | 4.2% | 0% |
| FrontierMath (Tiers 1-3) | 24.6% | — |
| FrontierMath Tier 4 | 0% | — |
| LiveBench Math | 90.2% | — |
Knowledge Gemini 2.5 Pro leads
Gemini 2.5 Pro: 56.0 (#46), Grok 3: 46.2 (#82)
| Benchmark | Gemini 2.5 Pro | Grok 3 |
|---|---|---|
| GPQA Diamond | 85.3% | 75.8% |
| MMLU-Pro | 86.3% | 78.8% |
| Confabulations | 10.6% | 14.2% |
| Vectara Hallucination Rate | 7% | 5.8% |
| GPQA (HELM) | 74.9% | 65% |
| LMArena Expert | 1452 | 1421 |
| Humanity's Last Exam | 21.6% | — |
Multimodal Not comparable
Gemini 2.5 Pro: 45.2 (#18), Grok 3: —
| Benchmark | Gemini 2.5 Pro | Grok 3 |
|---|---|---|
| LMArena Vision | 1263 | — |
| GeoBench | 86% | — |
| VPCT | 48% | — |
| LMArena Document | 1421 | — |
| SpatialViz-Bench | 44.7% | — |
Multilingual Gemini 2.5 Pro leads
Gemini 2.5 Pro: 55.3 (#31), Grok 3: 52.3 (#87)
| Benchmark | Gemini 2.5 Pro | Grok 3 |
|---|---|---|
| LMArena Non-English | 1451 | 1410 |
| LMArena Chinese | 1507 | 1448 |
| LMArena French | 1472 | 1460 |
| LMArena German | 1487 | 1431 |
| LMArena Japanese | 1461 | 1387 |
| LMArena Korean | 1434 | 1373 |
| LMArena Russian | 1461 | 1416 |
| LMArena Spanish | 1473 | 1417 |
Instruction Following Too close to call
Gemini 2.5 Pro: 75.0 (#75), Grok 3: 75.0 (#73)
| Benchmark | Gemini 2.5 Pro | Grok 3 |
|---|---|---|
| IFEval | 84% | 88.4% |
| LMArena Instruction Following | 1437 | 1409 |
| LiveBench Instruction Following | 80.6% | — |
Long Context Gemini 2.5 Pro leads
Gemini 2.5 Pro: 59.8 (#5), Grok 3: 38.7 (#192)
| Benchmark | Gemini 2.5 Pro | Grok 3 |
|---|---|---|
| Fiction.LiveBench | 91.7% | 58.3% |
| LMArena Longer Query | 1449 | 1439 |
Writing & Preference Gemini 2.5 Pro leads
Gemini 2.5 Pro: 63.7 (#62), Grok 3: 55.8 (#141)
| Benchmark | Gemini 2.5 Pro | Grok 3 |
|---|---|---|
| LMArena Text | 1458 | 1426 |
| LMArena Creative Writing | 1454 | 1414 |
| Short-Story Creative Writing | 83.8% | 76.4% |
| EQ-Bench Creative Writing | 1421 | 1186 |
| WildBench | 85.7% | 84.9% |
| LMArena Multi-Turn | 1453 | 1425 |
| LiveBench Language | 67.8% | — |
Frequently asked questions
Is Gemini 2.5 Pro better than Grok 3?
Gemini 2.5 Pro is the stronger model overall, scoring 45.0 to 39.9 on the Noometry Index.
Is Gemini 2.5 Pro or Grok 3 better for coding?
They score almost the same on coding (42.4 vs 41.9); test both on your own repository before choosing.
How many benchmarks do Gemini 2.5 Pro and Grok 3 share?
40 benchmarks have published results for both models. Gemini 2.5 Pro has 78 scored results on Noometry and Grok 3 has 40.