Model comparison
MiniMax-M3 vs Step 3.5 Flash
MiniMax-M3 is the stronger model overall, scoring 43.8 to 42.3 on the Noometry Index. Step 3.5 Flash costs 3.5× less per token, which makes it the better buy when MiniMax-M3's lead doesn't matter for your workload.
Last verified . 18 shared benchmarks.
MiniMax
StepFunSummary
- They share 18 benchmarks with published results for both. MiniMax-M3 scores higher in 6 categories and Step 3.5 Flash in 2 categories; 7 gaps are clear of the uncertainty.
- The widest gap is in knowledge, where MiniMax-M3 leads 58.4 to 39.6.
- The biggest single-benchmark swing is NYT Connections (extended): 65.1% for MiniMax-M3 and 28.4% for Step 3.5 Flash.
- Step 3.5 Flash is cheaper at $0.10 / $0.30 per million input/output tokens, against $0.30 / $1.20 for MiniMax-M3.
- MiniMax-M3 accepts more context: 1M tokens versus 256K.
Side by side
| MiniMax-M3 | Step 3.5 Flash | |
|---|---|---|
| Provider | MiniMax | StepFun |
| Noometry Index | 43.8 | 42.3 |
| Released | 2026-06-01 | 2026-01-29 |
| Weights | Open | Open |
| Context window | 1M | 256K |
| Max output | 512K | 256K |
| Input $ / M tokens | $0.30 | $0.10 |
| Output $ / M tokens | $1.20 | $0.30 |
| Results tracked | 41 | 19 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Too close to call
MiniMax-M3: 41.8 (#118), Step 3.5 Flash: 42.4 (#105)
| Benchmark | MiniMax-M3 | Step 3.5 Flash |
|---|---|---|
| LMArena Coding | 1469 | 1436 |
| FrontierCode | 14.7% | — |
| LMArena WebDev | 1482 | — |
| SciCode | 47.1% | — |
| ALE-Bench | 640.02 | — |
Agentic & Tool Use Not comparable
MiniMax-M3: 22.6 (#130), Step 3.5 Flash: —
| Benchmark | MiniMax-M3 | Step 3.5 Flash |
|---|---|---|
| APEX-Agents | 37.7% | — |
| OSWorld 2.0 | 4.6% | — |
| GBAEval | 0.9% | — |
| Vending-Bench 2 | 2,158 | — |
Reasoning MiniMax-M3 leads
MiniMax-M3: 30.1 (#87), Step 3.5 Flash: 22.2 (#202)
| Benchmark | MiniMax-M3 | Step 3.5 Flash |
|---|---|---|
| NYT Connections (extended) | 65.1% | 28.4% |
| LMArena Hard Prompts | 1447 | 1411 |
| SimpleBench | 45.8% | — |
| CritPt | 3.7% | — |
| Chess Puzzles | 14% | — |
| Mystery Game Puzzles | 8% | — |
| DTBench | 78.9% | — |
| LMCA | 33.7% | — |
| Surface Evolver Bench | 55% | — |
| Epoch Capabilities Index | 146.95 | — |
| ForecastBench | 61.4 | — |
Math Step 3.5 Flash leads
MiniMax-M3: 40.0 (#95), Step 3.5 Flash: 42.6 (#84)
| Benchmark | MiniMax-M3 | Step 3.5 Flash |
|---|---|---|
| LMArena Math | 1429 | 1408 |
| MathArena Final-Answer Competitions | — | 66.8% |
| OTIS Mock AIME 2024-2025 | 71.1% | — |
| ProofBench | 18% | — |
Knowledge MiniMax-M3 leads
MiniMax-M3: 58.4 (#35), Step 3.5 Flash: 39.6 (#132)
| Benchmark | MiniMax-M3 | Step 3.5 Flash |
|---|---|---|
| LMArena Expert | 1461 | 1421 |
| GPQA Diamond | 90.9% | — |
Multimodal Not comparable
MiniMax-M3: 40.2 (#51), Step 3.5 Flash: —
| Benchmark | MiniMax-M3 | Step 3.5 Flash |
|---|---|---|
| LMArena Vision | 1253 | — |
| LMArena Document | 1435 | — |
Multilingual MiniMax-M3 leads
MiniMax-M3: 53.0 (#75), Step 3.5 Flash: 50.5 (#119)
| Benchmark | MiniMax-M3 | Step 3.5 Flash |
|---|---|---|
| LMArena Non-English | 1420 | 1385 |
| LMArena Chinese | 1463 | 1447 |
| LMArena French | 1447 | 1421 |
| LMArena German | 1426 | 1405 |
| LMArena Japanese | 1381 | 1354 |
| LMArena Korean | 1372 | 1352 |
| LMArena Russian | 1428 | 1385 |
| LMArena Spanish | 1432 | 1419 |
Instruction Following MiniMax-M3 leads
MiniMax-M3: 75.5 (#62), Step 3.5 Flash: 73.1 (#124)
| Benchmark | MiniMax-M3 | Step 3.5 Flash |
|---|---|---|
| LMArena Instruction Following | 1433 | 1385 |
Long Context MiniMax-M3 leads
MiniMax-M3: 44.2 (#72), Step 3.5 Flash: 42.8 (#117)
| Benchmark | MiniMax-M3 | Step 3.5 Flash |
|---|---|---|
| LMArena Longer Query | 1445 | 1402 |
Writing & Preference MiniMax-M3 leads
MiniMax-M3: 62.1 (#83), Step 3.5 Flash: 58.8 (#113)
| Benchmark | MiniMax-M3 | Step 3.5 Flash |
|---|---|---|
| LMArena Text | 1433 | 1403 |
| LMArena Creative Writing | 1404 | 1357 |
| LMArena Multi-Turn | 1442 | 1405 |
| EQ-Bench 4 | 1150 | — |
Frequently asked questions
Is MiniMax-M3 better than Step 3.5 Flash?
MiniMax-M3 is the stronger model overall, scoring 43.8 to 42.3 on the Noometry Index. Step 3.5 Flash costs 3.5× less per token, which makes it the better buy when MiniMax-M3's lead doesn't matter for your workload.
Which is cheaper, MiniMax-M3 or Step 3.5 Flash?
Step 3.5 Flash is cheaper. It lists at $0.10 per million input tokens and $0.30 per million output tokens; MiniMax-M3 lists at $0.30 and $1.20.
Is MiniMax-M3 or Step 3.5 Flash better for coding?
They score almost the same on coding (41.8 vs 42.4); test both on your own repository before choosing.
Which has the bigger context window?
MiniMax-M3 does, with 1M tokens against 256K.
How many benchmarks do MiniMax-M3 and Step 3.5 Flash share?
18 benchmarks have published results for both models. MiniMax-M3 has 41 scored results on Noometry and Step 3.5 Flash has 19.