Model comparison
Gemma 2 2b IT vs Grok 4
Grok 4 is the stronger model overall, scoring 48.1 to 33.1 on the Noometry Index.
Last verified . 17 shared benchmarks.
GoogleSummary
- They share 17 benchmarks with published results for both. Gemma 2 2b IT scores higher in 0 categories and Grok 4 in 8 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in long context, where Grok 4 leads 63.1 to 34.3.
- Gemma 2 2b IT has downloadable open weights; the other is API-only.
Side by side
| Gemma 2 2b IT | Grok 4 | |
|---|---|---|
| Provider | xAI | |
| Noometry Index | 33.1 | 48.1 |
| Released | — | 2025-07-09 |
| Weights | Open | Proprietary |
| Context window | — | — |
| Max output | — | — |
| Input $ / M tokens | — | — |
| Output $ / M tokens | — | — |
| Results tracked | 17 | 48 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Grok 4 leads
Gemma 2 2b IT: 32.3 (#273), Grok 4: 50.3 (#46)
| Benchmark | Gemma 2 2b IT | Grok 4 |
|---|---|---|
| LMArena Coding | 1112 | 1408 |
| Aider Polyglot | — | 79.6% |
| WeirdML | — | 45.7% |
Agentic & Tool Use Not comparable
Gemma 2 2b IT: —, Grok 4: 32.3 (#68)
| Benchmark | Gemma 2 2b 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 2 2b IT: 21.4 (#224), Grok 4: 36.7 (#65)
| Benchmark | Gemma 2 2b IT | Grok 4 |
|---|---|---|
| LMArena Hard Prompts | 1113 | 1409 |
| ARC-AGI-2 | — | 16% |
| SimpleBench | — | 60.5% |
| Kagi LLM Benchmark | — | 73.6% |
| ARC-AGI-1 | — | 66.7% |
| Chess Puzzles | — | 28% |
| Epoch Capabilities Index | — | 146.44 |
| ForecastBench | — | 60.9 |
Math Grok 4 leads
Gemma 2 2b IT: 32.6 (#212), Grok 4: 48.4 (#64)
| Benchmark | Gemma 2 2b IT | Grok 4 |
|---|---|---|
| LMArena Math | 1135 | 1422 |
| OTIS Mock AIME 2024-2025 | — | 84% |
| Omni-MATH | — | 60.3% |
| FrontierMath (Feb 2025 set) | — | 19.7% |
| FrontierMath Tier 4 (v1) | — | 2.1% |
Knowledge Grok 4 leads
Gemma 2 2b IT: 29.9 (#231), Grok 4: 53.8 (#55)
| Benchmark | Gemma 2 2b IT | Grok 4 |
|---|---|---|
| LMArena Expert | 1096 | 1415 |
| GPQA Diamond | — | 87% |
| MMLU-Pro | — | 85.1% |
| Confabulations | — | 12.4% |
| GPQA (HELM) | — | 72.7% |
Multimodal Not comparable
Gemma 2 2b IT: —, Grok 4: 33.7 (#94)
| Benchmark | Gemma 2 2b IT | Grok 4 |
|---|---|---|
| LMArena Vision | — | 1210 |
| GeoBench | — | 45% |
Multilingual Grok 4 leads
Gemma 2 2b IT: 32.3 (#257), Grok 4: 51.8 (#103)
| Benchmark | Gemma 2 2b IT | Grok 4 |
|---|---|---|
| LMArena Non-English | 1121 | 1403 |
| LMArena Chinese | 1132 | 1427 |
| LMArena French | 1157 | 1418 |
| LMArena German | 1114 | 1429 |
| LMArena Japanese | 1083 | 1394 |
| LMArena Korean | 1055 | 1377 |
| LMArena Russian | 1118 | 1410 |
| LMArena Spanish | 1140 | 1420 |
Instruction Following Grok 4 leads
Gemma 2 2b IT: 57.8 (#263), Grok 4: 79.2 (#5)
| Benchmark | Gemma 2 2b IT | Grok 4 |
|---|---|---|
| LMArena Instruction Following | 1118 | 1387 |
| IFEval | — | 94.9% |
Long Context Grok 4 leads
Gemma 2 2b IT: 34.3 (#250), Grok 4: 63.1 (#4)
| Benchmark | Gemma 2 2b IT | Grok 4 |
|---|---|---|
| LMArena Longer Query | 1130 | 1409 |
| Fiction.LiveBench | — | 94.4% |
Writing & Preference Grok 4 leads
Gemma 2 2b IT: 36.5 (#263), Grok 4: 58.5 (#116)
| Benchmark | Gemma 2 2b IT | Grok 4 |
|---|---|---|
| LMArena Text | 1156 | 1411 |
| LMArena Creative Writing | 1147 | 1397 |
| LMArena Multi-Turn | 1118 | 1416 |
| Short-Story Creative Writing | — | 76.9% |
| WildBench | — | 79.7% |
Frequently asked questions
Is Gemma 2 2b IT better than Grok 4?
Grok 4 is the stronger model overall, scoring 48.1 to 33.1 on the Noometry Index.
Is Gemma 2 2b IT or Grok 4 better for coding?
Grok 4 scores higher on coding benchmarks: 50.3 versus 32.3 in the Noometry coding category.
How many benchmarks do Gemma 2 2b IT and Grok 4 share?
17 benchmarks have published results for both models. Gemma 2 2b IT has 17 scored results on Noometry and Grok 4 has 48.