Model comparison
Gemma 4 31B IT vs Grok 4
Grok 4 is the stronger model overall, scoring 48.1 to 43.5 on the Noometry Index.
Last verified . 21 shared benchmarks.
GoogleSummary
- They share 21 benchmarks with published results for both. Gemma 4 31B IT scores higher in 3 categories and Grok 4 in 6 categories; 9 gaps are clear of the uncertainty.
- The widest gap is in long context, where Grok 4 leads 63.1 to 44.2.
- The biggest single-benchmark swing is Chess Puzzles: 5% for Gemma 4 31B IT and 28% for Grok 4.
- Gemma 4 31B IT has downloadable open weights; the other is API-only.
Side by side
| Gemma 4 31B IT | Grok 4 | |
|---|---|---|
| Provider | xAI | |
| Noometry Index | 43.5 | 48.1 |
| Released | 2026-04-02 | 2025-07-09 |
| Weights | Open | Proprietary |
| Context window | 262K | — |
| Max output | 33K | — |
| Input $ / M tokens | $0.09 | — |
| Output $ / M tokens | $0.34 | — |
| Results tracked | 35 | 48 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Grok 4 leads
Gemma 4 31B IT: 42.3 (#108), Grok 4: 50.3 (#46)
| Benchmark | Gemma 4 31B IT | Grok 4 |
|---|---|---|
| WeirdML | 52.3% | 45.7% |
| LMArena Coding | 1459 | 1408 |
| Aider Polyglot | — | 79.6% |
| LMArena WebDev | 1366 | — |
| SciCode | 43.4% | — |
| ALE-Bench | 925.5 | — |
Agentic & Tool Use Not comparable
Gemma 4 31B IT: —, Grok 4: 32.3 (#68)
| Benchmark | Gemma 4 31B IT | Grok 4 |
|---|---|---|
| Terminal-Bench | — | 27.2% |
| Berkeley Function Calling Leaderboard | — | 63% |
| GDPval | — | 21.1% |
| Cybench | — | 43% |
| DeepResearch Bench | — | 47.3% |
| BALROG | — | 43.6% |
| LMArena Search | — | 1142 |
| METR Time Horizons | — | 66.6% |
Reasoning Grok 4 leads
Gemma 4 31B IT: 27.2 (#122), Grok 4: 36.7 (#65)
| Benchmark | Gemma 4 31B IT | Grok 4 |
|---|---|---|
| Kagi LLM Benchmark | 63.5% | 73.6% |
| Chess Puzzles | 5% | 28% |
| LMArena Hard Prompts | 1448 | 1409 |
| Epoch Capabilities Index | 142.74 | 146.44 |
| ARC-AGI-2 | — | 16% |
| SimpleBench | — | 60.5% |
| NYT Connections (extended) | 70.6% | — |
| ARC-AGI-1 | — | 66.7% |
| CritPt | 1.4% | — |
| Thematic Generalization | 53% | — |
| DTBench | 82.7% | — |
| LMCA | 39.3% | — |
| Surface Evolver Bench | 30.6% | — |
| ForecastBench | — | 60.9 |
Math Grok 4 leads
Gemma 4 31B IT: 43.2 (#81), Grok 4: 48.4 (#64)
| Benchmark | Gemma 4 31B IT | Grok 4 |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 73.3% | 84% |
| LMArena Math | 1465 | 1422 |
| Omni-MATH | — | 60.3% |
| FrontierMath (Feb 2025 set) | — | 19.7% |
| FrontierMath Tier 4 (v1) | — | 2.1% |
Knowledge Grok 4 leads
Gemma 4 31B IT: 37.9 (#151), Grok 4: 53.8 (#55)
| Benchmark | Gemma 4 31B IT | Grok 4 |
|---|---|---|
| GPQA Diamond | 75.8% | 87% |
| LMArena Expert | 1465 | 1415 |
| SimpleQA Verified | 10.4% | — |
| MMLU-Pro | — | 85.1% |
| Confabulations | — | 12.4% |
| Vectara Hallucination Rate | 7.4% | — |
| GPQA (HELM) | — | 72.7% |
Multimodal Gemma 4 31B IT leads
Gemma 4 31B IT: 41.6 (#34), Grok 4: 33.7 (#94)
| Benchmark | Gemma 4 31B IT | Grok 4 |
|---|---|---|
| LMArena Vision | 1277 | 1210 |
| GeoBench | — | 45% |
| LMArena Document | 1425 | — |
Multilingual Gemma 4 31B IT leads
Gemma 4 31B IT: 53.8 (#57), Grok 4: 51.8 (#103)
| Benchmark | Gemma 4 31B IT | Grok 4 |
|---|---|---|
| LMArena Non-English | 1431 | 1403 |
| LMArena Chinese | 1476 | 1427 |
| LMArena French | 1435 | 1418 |
| LMArena Russian | 1460 | 1410 |
| LMArena Spanish | 1444 | 1420 |
| LMArena German | — | 1429 |
| LMArena Japanese | — | 1394 |
| LMArena Korean | — | 1377 |
Instruction Following Grok 4 leads
Gemma 4 31B IT: 75.5 (#61), Grok 4: 79.2 (#5)
| Benchmark | Gemma 4 31B IT | Grok 4 |
|---|---|---|
| LMArena Instruction Following | 1433 | 1387 |
| IFEval | — | 94.9% |
Long Context Grok 4 leads
Gemma 4 31B IT: 44.2 (#71), Grok 4: 63.1 (#4)
| Benchmark | Gemma 4 31B IT | Grok 4 |
|---|---|---|
| LMArena Longer Query | 1446 | 1409 |
| Fiction.LiveBench | — | 94.4% |
Writing & Preference Gemma 4 31B IT leads
Gemma 4 31B IT: 60.5 (#96), Grok 4: 58.5 (#116)
| Benchmark | Gemma 4 31B IT | Grok 4 |
|---|---|---|
| LMArena Text | 1443 | 1411 |
| LMArena Creative Writing | 1415 | 1397 |
| LMArena Multi-Turn | 1452 | 1416 |
| Short-Story Creative Writing | — | 76.9% |
| EQ-Bench Creative Writing | 1368 | — |
| WildBench | — | 79.7% |
| EQ-Bench 4 | 1120 | — |
Frequently asked questions
Is Gemma 4 31B IT better than Grok 4?
Grok 4 is the stronger model overall, scoring 48.1 to 43.5 on the Noometry Index.
Is Gemma 4 31B IT or Grok 4 better for coding?
Grok 4 scores higher on coding benchmarks: 50.3 versus 42.3 in the Noometry coding category.
How many benchmarks do Gemma 4 31B IT and Grok 4 share?
21 benchmarks have published results for both models. Gemma 4 31B IT has 35 scored results on Noometry and Grok 4 has 48.