Model comparison
MiniMax-M2.5 vs Step 3
Step 3 is the stronger model overall, scoring 40.5 to 38.3 on the Noometry Index.
Last verified . 16 shared benchmarks.
MiniMax
StepFunSummary
- They share 16 benchmarks with published results for both. MiniMax-M2.5 scores higher in 4 categories and Step 3 in 4 categories; 6 gaps are clear of the uncertainty.
- The widest gap is in reasoning, where Step 3 leads 28.4 to 17.5.
- The biggest single-benchmark swing is Kagi LLM Benchmark: 55.2% for MiniMax-M2.5 and 62.3% for Step 3.
Side by side
| MiniMax-M2.5 | Step 3 | |
|---|---|---|
| Provider | MiniMax | StepFun |
| Noometry Index | 38.3 | 40.5 |
| Released | 2026-02-12 | — |
| Weights | Open | Open |
| Context window | 205K | — |
| Max output | 131K | — |
| Input $ / M tokens | $0.30 | — |
| Output $ / M tokens | $1.20 | — |
| Results tracked | 33 | 17 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding MiniMax-M2.5 leads
MiniMax-M2.5: 48.1 (#58), Step 3: 40.1 (#147)
| Benchmark | MiniMax-M2.5 | Step 3 |
|---|---|---|
| LMArena Coding | 1381 | 1367 |
| SWE-bench Verified (bash only) | 75.8% | — |
| LMArena WebDev | 1387 | — |
| SWE-bench Multilingual | 68.3% | — |
| ALE-Bench | 618.17 | — |
Agentic & Tool Use Not comparable
MiniMax-M2.5: 30.4 (#77), Step 3: —
| Benchmark | MiniMax-M2.5 | Step 3 |
|---|---|---|
| Terminal-Bench | 42.7% | — |
| Vending-Bench 2 | -23.16 | — |
Reasoning Step 3 leads
MiniMax-M2.5: 17.5 (#292), Step 3: 28.4 (#105)
| Benchmark | MiniMax-M2.5 | Step 3 |
|---|---|---|
| Kagi LLM Benchmark | 55.2% | 62.3% |
| LMArena Hard Prompts | 1372 | 1355 |
| ARC-AGI-2 | 4.9% | — |
| NYT Connections (extended) | 16.8% | — |
| ARC-AGI-1 | 63.7% | — |
| Epoch Capabilities Index | 146.68 | — |
Math Step 3 leads
MiniMax-M2.5: 26.9 (#253), Step 3: 37.6 (#148)
| Benchmark | MiniMax-M2.5 | Step 3 |
|---|---|---|
| LMArena Math | 1378 | 1366 |
| ProofBench | 4% | — |
Knowledge MiniMax-M2.5 leads
MiniMax-M2.5: 39.2 (#135), Step 3: 36.8 (#164)
| Benchmark | MiniMax-M2.5 | Step 3 |
|---|---|---|
| LMArena Expert | 1379 | 1333 |
| Vectara Hallucination Rate | 9.1% | — |
Multimodal Not comparable
MiniMax-M2.5: —, Step 3: 35.5 (#86)
| Benchmark | MiniMax-M2.5 | Step 3 |
|---|---|---|
| LMArena Vision | — | 1177 |
Multilingual Too close to call
MiniMax-M2.5: 47.1 (#152), Step 3: 46.3 (#159)
| Benchmark | MiniMax-M2.5 | Step 3 |
|---|---|---|
| LMArena Non-English | 1338 | 1327 |
| LMArena Chinese | 1393 | 1397 |
| LMArena German | 1362 | 1371 |
| LMArena Korean | 1232 | 1269 |
| LMArena Russian | 1358 | 1331 |
| LMArena Spanish | 1354 | 1371 |
| LMArena French | 1362 | — |
| LMArena Japanese | 1171 | — |
Instruction Following MiniMax-M2.5 leads
MiniMax-M2.5: 71.5 (#148), Step 3: 70.4 (#164)
| Benchmark | MiniMax-M2.5 | Step 3 |
|---|---|---|
| LMArena Instruction Following | 1353 | 1332 |
Long Context Step 3 leads
MiniMax-M2.5: 37.5 (#216), Step 3: 40.3 (#157)
| Benchmark | MiniMax-M2.5 | Step 3 |
|---|---|---|
| LMArena Longer Query | 1366 | 1326 |
| CL-bench | 11.4% | — |
| CL-bench Life | 6.3% | — |
Writing & Preference Too close to call
MiniMax-M2.5: 53.9 (#153), Step 3: 54.3 (#151)
| Benchmark | MiniMax-M2.5 | Step 3 |
|---|---|---|
| LMArena Text | 1359 | 1350 |
| LMArena Creative Writing | 1331 | 1321 |
| LMArena Multi-Turn | 1364 | 1341 |
| EQ-Bench Creative Writing | 1361 | — |
Frequently asked questions
Is MiniMax-M2.5 better than Step 3?
Step 3 is the stronger model overall, scoring 40.5 to 38.3 on the Noometry Index.
Is MiniMax-M2.5 or Step 3 better for coding?
MiniMax-M2.5 scores higher on coding benchmarks: 48.1 versus 40.1 in the Noometry coding category.
How many benchmarks do MiniMax-M2.5 and Step 3 share?
16 benchmarks have published results for both models. MiniMax-M2.5 has 33 scored results on Noometry and Step 3 has 17.