Model comparison
Llama 3-70B vs MiniMax-M2
MiniMax-M2 is the stronger model overall, scoring 37.4 to 28.8 on the Noometry Index.
Last verified . 16 shared benchmarks.
Meta
MiniMaxSummary
- They share 16 benchmarks with published results for both. Llama 3-70B scores higher in 0 categories and MiniMax-M2 in 9 categories; 9 gaps are clear of the uncertainty.
- The widest gap is in math, where MiniMax-M2 leads 37.3 to 12.8.
- The biggest single-benchmark swing is Kagi LLM Benchmark: 35.1% for Llama 3-70B and 57.8% for MiniMax-M2.
Side by side
| Llama 3-70B | MiniMax-M2 | |
|---|---|---|
| Provider | Meta | MiniMax |
| Noometry Index | 28.8 | 37.4 |
| Released | 2024-04-18 | 2025-10-27 |
| Weights | Open | Open |
| Context window | — | 205K |
| Max output | — | 131K |
| Input $ / M tokens | — | $0.30 |
| Output $ / M tokens | — | $1.20 |
| Results tracked | 31 | 21 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding MiniMax-M2 leads
Llama 3-70B: 35.8 (#218), MiniMax-M2: 39.3 (#159)
| Benchmark | Llama 3-70B | MiniMax-M2 |
|---|---|---|
| LMArena Coding | 1206 | 1370 |
| SWE-bench Verified (bash only) | — | 61% |
| LMArena WebDev | — | 1297 |
| BigCodeBench Instruct | 43.6% | — |
| BigCodeBench Complete | 54.5% | — |
| HumanEval+ | 72% | — |
| MBPP+ | 69% | — |
Agentic & Tool Use MiniMax-M2 leads
Llama 3-70B: 21.1 (#139), MiniMax-M2: 25.1 (#109)
| Benchmark | Llama 3-70B | MiniMax-M2 |
|---|---|---|
| Terminal-Bench | — | 30% |
| Cybench | 5% | — |
| Vending-Bench 2 | — | 160.6 |
Reasoning MiniMax-M2 leads
Llama 3-70B: 18.0 (#288), MiniMax-M2: 19.4 (#258)
| Benchmark | Llama 3-70B | MiniMax-M2 |
|---|---|---|
| Kagi LLM Benchmark | 35.1% | 57.8% |
| LMArena Hard Prompts | 1195 | 1357 |
| NYT Connections (extended) | — | 14.8% |
| DTBench | 54.2% | — |
| Epoch Capabilities Index | 122.93 | — |
| ForecastBench | 57.1 | — |
| WinoGrande | 83.5% | — |
Math MiniMax-M2 leads
Llama 3-70B: 12.8 (#305), MiniMax-M2: 37.3 (#160)
| Benchmark | Llama 3-70B | MiniMax-M2 |
|---|---|---|
| LMArena Math | 1218 | 1352 |
| OTIS Mock AIME 2024-2025 | 4.3% | — |
| MATH Level 5 | 22.6% | — |
Knowledge MiniMax-M2 leads
Llama 3-70B: 20.8 (#277), MiniMax-M2: 37.0 (#163)
| Benchmark | Llama 3-70B | MiniMax-M2 |
|---|---|---|
| LMArena Expert | 1149 | 1337 |
| GPQA Diamond | 40.6% | — |
| MMLU | 79.3% | — |
Multilingual MiniMax-M2 leads
Llama 3-70B: 33.6 (#251), MiniMax-M2: 45.3 (#171)
| Benchmark | Llama 3-70B | MiniMax-M2 |
|---|---|---|
| LMArena Non-English | 1142 | 1313 |
| LMArena Chinese | 1114 | 1366 |
| LMArena French | 1232 | 1335 |
| LMArena German | 1169 | 1355 |
| LMArena Russian | 1159 | 1331 |
| LMArena Spanish | 1241 | 1326 |
| LMArena Japanese | 1017 | — |
| LMArena Korean | 1017 | — |
Instruction Following MiniMax-M2 leads
Llama 3-70B: 62.5 (#238), MiniMax-M2: 70.2 (#166)
| Benchmark | Llama 3-70B | MiniMax-M2 |
|---|---|---|
| LMArena Instruction Following | 1194 | 1328 |
Long Context MiniMax-M2 leads
Llama 3-70B: 35.6 (#240), MiniMax-M2: 40.5 (#153)
| Benchmark | Llama 3-70B | MiniMax-M2 |
|---|---|---|
| LMArena Longer Query | 1174 | 1331 |
Writing & Preference MiniMax-M2 leads
Llama 3-70B: 42.8 (#231), MiniMax-M2: 53.0 (#162)
| Benchmark | Llama 3-70B | MiniMax-M2 |
|---|---|---|
| LMArena Text | 1221 | 1340 |
| LMArena Creative Writing | 1210 | 1286 |
| LMArena Multi-Turn | 1223 | 1361 |
Frequently asked questions
Is Llama 3-70B better than MiniMax-M2?
MiniMax-M2 is the stronger model overall, scoring 37.4 to 28.8 on the Noometry Index.
Is Llama 3-70B or MiniMax-M2 better for coding?
MiniMax-M2 scores higher on coding benchmarks: 39.3 versus 35.8 in the Noometry coding category.
How many benchmarks do Llama 3-70B and MiniMax-M2 share?
16 benchmarks have published results for both models. Llama 3-70B has 31 scored results on Noometry and MiniMax-M2 has 21.