Model comparison
Llama 3.1-405B vs MiniMax-M2.5
MiniMax-M2.5 is the stronger model overall, scoring 38.3 to 30.7 on the Noometry Index.
Last verified . 20 shared benchmarks.
Meta
MiniMaxSummary
- They share 20 benchmarks with published results for both. Llama 3.1-405B scores higher in 1 category and MiniMax-M2.5 in 8 categories; 7 gaps are clear of the uncertainty.
- The widest gap is in writing & preference, where MiniMax-M2.5 leads 53.9 to 38.9.
- The biggest single-benchmark swing is Kagi LLM Benchmark: 45% for Llama 3.1-405B and 55.2% for MiniMax-M2.5.
Side by side
| Llama 3.1-405B | MiniMax-M2.5 | |
|---|---|---|
| Provider | Meta | MiniMax |
| Noometry Index | 30.7 | 38.3 |
| Released | 2024-07-23 | 2026-02-12 |
| Weights | Open | Open |
| Context window | — | 205K |
| Max output | — | 131K |
| Input $ / M tokens | — | $0.30 |
| Output $ / M tokens | — | $1.20 |
| Results tracked | 42 | 33 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding MiniMax-M2.5 leads
Llama 3.1-405B: 33.1 (#262), MiniMax-M2.5: 48.1 (#58)
| Benchmark | Llama 3.1-405B | MiniMax-M2.5 |
|---|---|---|
| LMArena Coding | 1291 | 1381 |
| SWE-bench Verified (bash only) | — | 75.8% |
| LMArena WebDev | — | 1387 |
| SWE-bench Multilingual | — | 68.3% |
| WeirdML | 21.4% | — |
| ALE-Bench | — | 618.17 |
Agentic & Tool Use MiniMax-M2.5 leads
Llama 3.1-405B: 21.0 (#140), MiniMax-M2.5: 30.4 (#77)
| Benchmark | Llama 3.1-405B | MiniMax-M2.5 |
|---|---|---|
| Terminal-Bench | — | 42.7% |
| TheAgentCompany | 7.4% | — |
| Cybench | 7.5% | — |
| Vending-Bench 2 | — | -23.16 |
Reasoning Too close to call
Llama 3.1-405B: 16.8 (#300), MiniMax-M2.5: 17.5 (#292)
| Benchmark | Llama 3.1-405B | MiniMax-M2.5 |
|---|---|---|
| Kagi LLM Benchmark | 45% | 55.2% |
| LMArena Hard Prompts | 1269 | 1372 |
| Epoch Capabilities Index | 128.75 | 146.68 |
| ARC-AGI-2 | — | 4.9% |
| SimpleBench | 23% | — |
| NYT Connections (extended) | — | 16.8% |
| ARC-AGI-1 | — | 63.7% |
| DTBench | 61.4% | — |
| BIG-Bench Hard | 82.9% | — |
| ForecastBench | 59.9 | — |
| HellaSwag | 89.2% | — |
| PIQA | 85.9% | — |
| WinoGrande | 89.2% | — |
Math MiniMax-M2.5 leads
Llama 3.1-405B: 18.4 (#290), MiniMax-M2.5: 26.9 (#253)
| Benchmark | Llama 3.1-405B | MiniMax-M2.5 |
|---|---|---|
| LMArena Math | 1281 | 1378 |
| OTIS Mock AIME 2024-2025 | 9.7% | — |
| ProofBench | — | 4% |
| Omni-MATH | 24.9% | — |
| MATH Level 5 | 49.8% | — |
Knowledge MiniMax-M2.5 leads
Llama 3.1-405B: 30.4 (#227), MiniMax-M2.5: 39.2 (#135)
| Benchmark | Llama 3.1-405B | MiniMax-M2.5 |
|---|---|---|
| LMArena Expert | 1243 | 1379 |
| GPQA Diamond | 50.9% | — |
| MMLU-Pro | 72.3% | — |
| Confabulations | 17.6% | — |
| Vectara Hallucination Rate | — | 9.1% |
| GPQA (HELM) | 52.2% | — |
| ARC (AI2) Challenge | 95.3% | — |
| MMLU | 84.5% | — |
| TriviaQA | 82.7% | — |
Multilingual MiniMax-M2.5 leads
Llama 3.1-405B: 40.7 (#214), MiniMax-M2.5: 47.1 (#152)
| Benchmark | Llama 3.1-405B | MiniMax-M2.5 |
|---|---|---|
| LMArena Non-English | 1248 | 1338 |
| LMArena Chinese | 1242 | 1393 |
| LMArena French | 1279 | 1362 |
| LMArena German | 1252 | 1362 |
| LMArena Japanese | 1208 | 1171 |
| LMArena Korean | 1184 | 1232 |
| LMArena Russian | 1265 | 1358 |
| LMArena Spanish | 1260 | 1354 |
Instruction Following MiniMax-M2.5 leads
Llama 3.1-405B: 65.9 (#214), MiniMax-M2.5: 71.5 (#148)
| Benchmark | Llama 3.1-405B | MiniMax-M2.5 |
|---|---|---|
| LMArena Instruction Following | 1259 | 1353 |
| IFEval | 81.1% | — |
Long Context Too close to call
Llama 3.1-405B: 38.4 (#197), MiniMax-M2.5: 37.5 (#216)
| Benchmark | Llama 3.1-405B | MiniMax-M2.5 |
|---|---|---|
| LMArena Longer Query | 1266 | 1366 |
| CL-bench | — | 11.4% |
| CL-bench Life | — | 6.3% |
Writing & Preference MiniMax-M2.5 leads
Llama 3.1-405B: 38.9 (#251), MiniMax-M2.5: 53.9 (#153)
| Benchmark | Llama 3.1-405B | MiniMax-M2.5 |
|---|---|---|
| LMArena Text | 1284 | 1359 |
| LMArena Creative Writing | 1262 | 1331 |
| EQ-Bench Creative Writing | 870 | 1361 |
| LMArena Multi-Turn | 1297 | 1364 |
| WildBench | 78.3% | — |
Frequently asked questions
Is Llama 3.1-405B better than MiniMax-M2.5?
MiniMax-M2.5 is the stronger model overall, scoring 38.3 to 30.7 on the Noometry Index.
Is Llama 3.1-405B or MiniMax-M2.5 better for coding?
MiniMax-M2.5 scores higher on coding benchmarks: 48.1 versus 33.1 in the Noometry coding category.
How many benchmarks do Llama 3.1-405B and MiniMax-M2.5 share?
20 benchmarks have published results for both models. Llama 3.1-405B has 42 scored results on Noometry and MiniMax-M2.5 has 33.