Model comparison
Gemma 2 27B vs MiniMax-M3
MiniMax-M3 is the stronger model overall, scoring 43.8 to 29.4 on the Noometry Index.
Last verified . 22 shared benchmarks.
Google
MiniMaxSummary
- They share 22 benchmarks with published results for both. Gemma 2 27B scores higher in 0 categories and MiniMax-M3 in 8 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in knowledge, where MiniMax-M3 leads 58.4 to 19.0.
- The biggest single-benchmark swing is OTIS Mock AIME 2024-2025: 1.4% for Gemma 2 27B and 71.1% for MiniMax-M3.
- MiniMax-M3 is cheaper at $0.30 / $1.20 per million input/output tokens, against $0.65 / $0.65 for Gemma 2 27B.
- MiniMax-M3 accepts more context: 1M tokens versus 8K.
Side by side
| Gemma 2 27B | MiniMax-M3 | |
|---|---|---|
| Provider | MiniMax | |
| Noometry Index | 29.4 | 43.8 |
| Released | 2024-06-24 | 2026-06-01 |
| Weights | Open | Open |
| Context window | 8K | 1M |
| Max output | 2K | 512K |
| Input $ / M tokens | $0.65 | $0.30 |
| Output $ / M tokens | $0.65 | $1.20 |
| Results tracked | 34 | 41 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding MiniMax-M3 leads
Gemma 2 27B: 34.1 (#246), MiniMax-M3: 41.8 (#118)
| Benchmark | Gemma 2 27B | MiniMax-M3 |
|---|---|---|
| LMArena Coding | 1211 | 1469 |
| FrontierCode | — | 14.7% |
| LMArena WebDev | — | 1482 |
| SciCode | — | 47.1% |
| BigCodeBench Instruct | 42.8% | — |
| LiveBench Coding | 36% | — |
| BigCodeBench Complete | 52.5% | — |
| ALE-Bench | — | 640.02 |
Agentic & Tool Use Not comparable
Gemma 2 27B: —, MiniMax-M3: 22.6 (#130)
| Benchmark | Gemma 2 27B | MiniMax-M3 |
|---|---|---|
| APEX-Agents | — | 37.7% |
| OSWorld 2.0 | — | 4.6% |
| GBAEval | — | 0.9% |
| Vending-Bench 2 | — | 2,158 |
Reasoning MiniMax-M3 leads
Gemma 2 27B: 15.3 (#315), MiniMax-M3: 30.1 (#87)
| Benchmark | Gemma 2 27B | MiniMax-M3 |
|---|---|---|
| LMArena Hard Prompts | 1198 | 1447 |
| DTBench | 48% | 78.9% |
| LMCA | 7.1% | 33.7% |
| Epoch Capabilities Index | 122.08 | 146.95 |
| SimpleBench | — | 45.8% |
| NYT Connections (extended) | — | 65.1% |
| CritPt | — | 3.7% |
| Chess Puzzles | — | 14% |
| LiveBench Reasoning | 28.1% | — |
| Mystery Game Puzzles | — | 8% |
| LiveBench Data Analysis | 47.9% | — |
| Surface Evolver Bench | — | 55% |
| ForecastBench | — | 61.4 |
| LiveBench | 38.2% | — |
Math MiniMax-M3 leads
Gemma 2 27B: 10.7 (#311), MiniMax-M3: 40.0 (#95)
| Benchmark | Gemma 2 27B | MiniMax-M3 |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 1.4% | 71.1% |
| LMArena Math | 1212 | 1429 |
| ProofBench | — | 18% |
| LiveBench Math | 26.5% | — |
| MATH Level 5 | 27.9% | — |
Knowledge MiniMax-M3 leads
Gemma 2 27B: 19.0 (#280), MiniMax-M3: 58.4 (#35)
| Benchmark | Gemma 2 27B | MiniMax-M3 |
|---|---|---|
| GPQA Diamond | 36.5% | 90.9% |
| LMArena Expert | 1172 | 1461 |
| Confabulations | 27.1% | — |
| MMLU | 75.7% | — |
Multimodal Not comparable
Gemma 2 27B: —, MiniMax-M3: 40.2 (#51)
| Benchmark | Gemma 2 27B | MiniMax-M3 |
|---|---|---|
| LMArena Vision | — | 1253 |
| LMArena Document | — | 1435 |
Multilingual MiniMax-M3 leads
Gemma 2 27B: 38.6 (#226), MiniMax-M3: 53.0 (#75)
| Benchmark | Gemma 2 27B | MiniMax-M3 |
|---|---|---|
| LMArena Non-English | 1217 | 1420 |
| LMArena Chinese | 1221 | 1463 |
| LMArena French | 1247 | 1447 |
| LMArena German | 1209 | 1426 |
| LMArena Japanese | 1175 | 1381 |
| LMArena Korean | 1174 | 1372 |
| LMArena Russian | 1234 | 1428 |
| LMArena Spanish | 1228 | 1432 |
Instruction Following MiniMax-M3 leads
Gemma 2 27B: 60.5 (#249), MiniMax-M3: 75.5 (#62)
| Benchmark | Gemma 2 27B | MiniMax-M3 |
|---|---|---|
| LMArena Instruction Following | 1206 | 1433 |
| LiveBench Instruction Following | 58.1% | — |
Long Context MiniMax-M3 leads
Gemma 2 27B: 37.3 (#218), MiniMax-M3: 44.2 (#72)
| Benchmark | Gemma 2 27B | MiniMax-M3 |
|---|---|---|
| LMArena Longer Query | 1231 | 1445 |
Writing & Preference MiniMax-M3 leads
Gemma 2 27B: 44.2 (#225), MiniMax-M3: 62.1 (#83)
| Benchmark | Gemma 2 27B | MiniMax-M3 |
|---|---|---|
| LMArena Text | 1231 | 1433 |
| LMArena Creative Writing | 1241 | 1404 |
| LMArena Multi-Turn | 1224 | 1442 |
| EQ-Bench 4 | — | 1150 |
| LiveBench Language | 32.6% | — |
Frequently asked questions
Is Gemma 2 27B better than MiniMax-M3?
MiniMax-M3 is the stronger model overall, scoring 43.8 to 29.4 on the Noometry Index.
Which is cheaper, Gemma 2 27B or MiniMax-M3?
MiniMax-M3 is cheaper. It lists at $0.30 per million input tokens and $1.20 per million output tokens; Gemma 2 27B lists at $0.65 and $0.65.
Is Gemma 2 27B or MiniMax-M3 better for coding?
MiniMax-M3 scores higher on coding benchmarks: 41.8 versus 34.1 in the Noometry coding category.
Which has the bigger context window?
MiniMax-M3 does, with 1M tokens against 8K.
How many benchmarks do Gemma 2 27B and MiniMax-M3 share?
22 benchmarks have published results for both models. Gemma 2 27B has 34 scored results on Noometry and MiniMax-M3 has 41.