Model comparison
Gemma 3 27B vs Grok-3 mini
Grok-3 mini is the stronger model overall, scoring 41.2 to 30.8 on the Noometry Index.
Last verified . 26 shared benchmarks.
GoogleSummary
- They share 26 benchmarks with published results for both. Gemma 3 27B scores higher in 2 categories and Grok-3 mini in 6 categories; 7 gaps are clear of the uncertainty.
- The widest gap is in knowledge, where Grok-3 mini leads 46.4 to 25.5.
- The biggest single-benchmark swing is OTIS Mock AIME 2024-2025: 22.5% for Gemma 3 27B and 77.8% for Grok-3 mini.
- Gemma 3 27B has downloadable open weights; the other is API-only.
Side by side
| Gemma 3 27B | Grok-3 mini | |
|---|---|---|
| Provider | xAI | |
| Noometry Index | 30.8 | 41.2 |
| Released | 2025-03-11 | 2025-04-09 |
| Weights | Open | Proprietary |
| Context window | 131K | — |
| Max output | 8K | — |
| Input $ / M tokens | $0.08 | — |
| Output $ / M tokens | $0.16 | — |
| Results tracked | 43 | 35 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Grok-3 mini leads
Gemma 3 27B: 22.5 (#334), Grok-3 mini: 40.8 (#131)
| Benchmark | Gemma 3 27B | Grok-3 mini |
|---|---|---|
| Aider Polyglot | 4.9% | 49.3% |
| LMArena Coding | 1322 | 1379 |
| SciCode | 21.2% | — |
| WeirdML | — | 42.6% |
| LiveBench Coding | 39.9% | — |
Agentic & Tool Use Not comparable
Gemma 3 27B: 25.1 (#110), Grok-3 mini: —
| Benchmark | Gemma 3 27B | Grok-3 mini |
|---|---|---|
| Berkeley Function Calling Leaderboard | 29.5% | — |
Reasoning Gemma 3 27B leads
Gemma 3 27B: 16.7 (#301), Grok-3 mini: 13.6 (#334)
| Benchmark | Gemma 3 27B | Grok-3 mini |
|---|---|---|
| Kagi LLM Benchmark | 40.4% | 61.3% |
| LMArena Hard Prompts | 1340 | 1375 |
| Epoch Capabilities Index | 130.04 | 140.35 |
| ARC-AGI-2 | — | 0.4% |
| ARC-AGI-1 | — | 16.5% |
| CritPt | 0% | — |
| Chess Puzzles | 0% | — |
| LiveBench Reasoning | 43.8% | — |
| DTBench | 52.5% | — |
| LiveBench Data Analysis | 51.5% | — |
| LMCA | 12.3% | — |
| LiveBench | 50% | — |
Math Grok-3 mini leads
Gemma 3 27B: 25.9 (#265), Grok-3 mini: 42.1 (#85)
| Benchmark | Gemma 3 27B | Grok-3 mini |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 22.5% | 77.8% |
| LMArena Math | 1312 | 1386 |
| MATH Level 5 | 74% | 90.9% |
| Omni-MATH | — | 31.8% |
| LiveBench Math | 55.4% | — |
| FrontierMath (Feb 2025 set) | — | 5.9% |
Knowledge Grok-3 mini leads
Gemma 3 27B: 25.5 (#261), Grok-3 mini: 46.4 (#81)
| Benchmark | Gemma 3 27B | Grok-3 mini |
|---|---|---|
| GPQA Diamond | 47.7% | 76.3% |
| Confabulations | 40.3% | 10.8% |
| LMArena Expert | 1304 | 1395 |
| MMLU-Pro | — | 79.9% |
| Vectara Hallucination Rate | 7.4% | — |
| GPQA (HELM) | — | 67.5% |
Multimodal Not comparable
Gemma 3 27B: 32.6 (#100), Grok-3 mini: —
| Benchmark | Gemma 3 27B | Grok-3 mini |
|---|---|---|
| LMArena Vision | 1164 | — |
| GeoBench | 52% | — |
Multilingual Grok-3 mini leads
Gemma 3 27B: 46.9 (#155), Grok-3 mini: 48.1 (#145)
| Benchmark | Gemma 3 27B | Grok-3 mini |
|---|---|---|
| LMArena Non-English | 1334 | 1352 |
| LMArena Chinese | 1346 | 1387 |
| LMArena French | 1368 | 1357 |
| LMArena German | 1362 | 1349 |
| LMArena Japanese | 1287 | 1342 |
| LMArena Korean | 1308 | 1335 |
| LMArena Russian | 1349 | 1353 |
| LMArena Spanish | 1349 | 1381 |
Instruction Following Grok-3 mini leads
Gemma 3 27B: 70.6 (#160), Grok-3 mini: 78.5 (#9)
| Benchmark | Gemma 3 27B | Grok-3 mini |
|---|---|---|
| LMArena Instruction Following | 1321 | 1357 |
| LiveBench Instruction Following | 74.9% | — |
| IFEval | — | 95.1% |
Long Context Grok-3 mini leads
Gemma 3 27B: 27.6 (#293), Grok-3 mini: 41.0 (#147)
| Benchmark | Gemma 3 27B | Grok-3 mini |
|---|---|---|
| Fiction.LiveBench | 33.3% | 66.7% |
| LMArena Longer Query | 1333 | 1372 |
Writing & Preference Too close to call
Gemma 3 27B: 52.5 (#168), Grok-3 mini: 52.5 (#169)
| Benchmark | Gemma 3 27B | Grok-3 mini |
|---|---|---|
| LMArena Text | 1358 | 1370 |
| LMArena Creative Writing | 1346 | 1342 |
| Short-Story Creative Writing | 79.9% | 73.5% |
| LMArena Multi-Turn | 1345 | 1355 |
| EQ-Bench Creative Writing | 1266 | — |
| WildBench | — | 65.1% |
| LiveBench Language | 34.6% | — |
Frequently asked questions
Is Gemma 3 27B better than Grok-3 mini?
Grok-3 mini is the stronger model overall, scoring 41.2 to 30.8 on the Noometry Index.
Is Gemma 3 27B or Grok-3 mini better for coding?
Grok-3 mini scores higher on coding benchmarks: 40.8 versus 22.5 in the Noometry coding category.
How many benchmarks do Gemma 3 27B and Grok-3 mini share?
26 benchmarks have published results for both models. Gemma 3 27B has 43 scored results on Noometry and Grok-3 mini has 35.