Model comparison
Gemma 2 27B vs Grok 4.6
Grok 4.6 is the stronger model overall, scoring 56.9 to 29.4 on the Noometry Index. Gemma 2 27B costs 4.6× less per token, which makes it the better buy when Grok 4.6's lead doesn't matter for your workload.
Last verified . 22 shared benchmarks.
GoogleSummary
- They share 22 benchmarks with published results for both. Gemma 2 27B scores higher in 0 categories and Grok 4.6 in 8 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in math, where Grok 4.6 leads 67.0 to 10.7.
- The biggest single-benchmark swing is OTIS Mock AIME 2024-2025: 1.4% for Gemma 2 27B and 99.2% for Grok 4.6.
- Gemma 2 27B is cheaper at $0.65 / $0.65 per million input/output tokens, against $2 / $6 for Grok 4.6.
- Grok 4.6 accepts more context: 500K tokens versus 8K.
- Gemma 2 27B has downloadable open weights; the other is API-only.
Side by side
| Gemma 2 27B | Grok 4.6 | |
|---|---|---|
| Provider | xAI | |
| Noometry Index | 29.4 | 56.9 |
| Released | 2024-06-24 | 2026-08-12 |
| Weights | Open | Proprietary |
| Context window | 8K | 500K |
| Max output | 2K | 500K |
| Input $ / M tokens | $0.65 | $2 |
| Output $ / M tokens | $0.65 | $6 |
| Results tracked | 34 | 49 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Grok 4.6 leads
Gemma 2 27B: 34.1 (#246), Grok 4.6: 58.5 (#16)
| Benchmark | Gemma 2 27B | Grok 4.6 |
|---|---|---|
| LMArena Coding | 1211 | 1465 |
| DeepSWE | — | 67.5% |
| FrontierCode | — | 48% |
| CursorBench | — | 41.4% |
| LMArena WebDev | — | 1617 |
| FrontierSWE | — | 25.3% |
| SciCode | — | 56.5% |
| WeirdML | — | 67.3% |
| BigCodeBench Instruct | 42.8% | — |
| LiveBench Coding | 36% | — |
| BigCodeBench Complete | 52.5% | — |
| ALE-Bench | — | 1,508 |
Agentic & Tool Use Not comparable
Gemma 2 27B: —, Grok 4.6: 39.4 (#27)
| Benchmark | Gemma 2 27B | Grok 4.6 |
|---|---|---|
| APEX-Agents | — | 65.3% |
| GDP.pdf | — | 17.2% |
| Vending-Bench 2 | — | 9,047 |
Reasoning Grok 4.6 leads
Gemma 2 27B: 15.3 (#315), Grok 4.6: 61.4 (#20)
| Benchmark | Gemma 2 27B | Grok 4.6 |
|---|---|---|
| LMArena Hard Prompts | 1198 | 1447 |
| DTBench | 48% | 97.3% |
| LMCA | 7.1% | 48.5% |
| Epoch Capabilities Index | 122.08 | 156.44 |
| ARC-AGI-2 | — | 67.1% |
| SimpleBench | — | 75.9% |
| NYT Connections (extended) | — | 80% |
| ARC-AGI-1 | — | 87.5% |
| CritPt | — | 19.7% |
| Chess Puzzles | — | 40% |
| EBR-Bench | — | 30.5% |
| LiveBench Reasoning | 28.1% | — |
| Mystery Game Puzzles | — | 34% |
| LiveBench Data Analysis | 47.9% | — |
| LiveBench | 38.2% | — |
Math Grok 4.6 leads
Gemma 2 27B: 10.7 (#311), Grok 4.6: 67.0 (#24)
| Benchmark | Gemma 2 27B | Grok 4.6 |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 1.4% | 99.2% |
| LMArena Math | 1212 | 1423 |
| FrontierMath (Tiers 1-3) | — | 66% |
| FrontierMath Tier 4 | — | 31.7% |
| ProofBench | — | 51% |
| LiveBench Math | 26.5% | — |
| MATH Level 5 | 27.9% | — |
Knowledge Grok 4.6 leads
Gemma 2 27B: 19.0 (#280), Grok 4.6: 63.3 (#20)
| Benchmark | Gemma 2 27B | Grok 4.6 |
|---|---|---|
| GPQA Diamond | 36.5% | 94% |
| LMArena Expert | 1172 | 1467 |
| SimpleQA Verified | — | 49.3% |
| Confabulations | 27.1% | — |
| MMLU | 75.7% | — |
Multimodal Not comparable
Gemma 2 27B: —, Grok 4.6: 43.6 (#23)
| Benchmark | Gemma 2 27B | Grok 4.6 |
|---|---|---|
| LMArena Vision | — | 1263 |
| Blueprint-Bench 2 | — | 33.2% |
| Furniture Assembly | — | 40% |
| LMArena Document | — | 1452 |
Multilingual Grok 4.6 leads
Gemma 2 27B: 38.6 (#226), Grok 4.6: 53.0 (#74)
| Benchmark | Gemma 2 27B | Grok 4.6 |
|---|---|---|
| LMArena Non-English | 1217 | 1420 |
| LMArena Chinese | 1221 | 1480 |
| LMArena French | 1247 | 1461 |
| LMArena German | 1209 | 1431 |
| LMArena Japanese | 1175 | 1376 |
| LMArena Korean | 1174 | 1397 |
| LMArena Russian | 1234 | 1422 |
| LMArena Spanish | 1228 | 1404 |
Instruction Following Grok 4.6 leads
Gemma 2 27B: 60.5 (#249), Grok 4.6: 75.4 (#63)
| Benchmark | Gemma 2 27B | Grok 4.6 |
|---|---|---|
| LMArena Instruction Following | 1206 | 1431 |
| LiveBench Instruction Following | 58.1% | — |
Long Context Grok 4.6 leads
Gemma 2 27B: 37.3 (#218), Grok 4.6: 44.5 (#66)
| Benchmark | Gemma 2 27B | Grok 4.6 |
|---|---|---|
| LMArena Longer Query | 1231 | 1454 |
Writing & Preference Grok 4.6 leads
Gemma 2 27B: 44.2 (#225), Grok 4.6: 62.3 (#80)
| Benchmark | Gemma 2 27B | Grok 4.6 |
|---|---|---|
| LMArena Text | 1231 | 1428 |
| LMArena Creative Writing | 1241 | 1428 |
| LMArena Multi-Turn | 1224 | 1425 |
| LiveBench Language | 32.6% | — |
Frequently asked questions
Is Gemma 2 27B better than Grok 4.6?
Grok 4.6 is the stronger model overall, scoring 56.9 to 29.4 on the Noometry Index. Gemma 2 27B costs 4.6× less per token, which makes it the better buy when Grok 4.6's lead doesn't matter for your workload.
Which is cheaper, Gemma 2 27B or Grok 4.6?
Gemma 2 27B is cheaper. It lists at $0.65 per million input tokens and $0.65 per million output tokens; Grok 4.6 lists at $2 and $6.
Is Gemma 2 27B or Grok 4.6 better for coding?
Grok 4.6 scores higher on coding benchmarks: 58.5 versus 34.1 in the Noometry coding category.
Which has the bigger context window?
Grok 4.6 does, with 500K tokens against 8K.
How many benchmarks do Gemma 2 27B and Grok 4.6 share?
22 benchmarks have published results for both models. Gemma 2 27B has 34 scored results on Noometry and Grok 4.6 has 49.