Model comparison
Magistral Medium vs Step 5 Preview
Step 5 Preview is the stronger model overall, scoring 47.9 to 35.2 on the Noometry Index.
Last verified . 15 shared benchmarks.
Mistral AI
StepFunSummary
- They share 15 benchmarks with published results for both. Magistral Medium scores higher in 0 categories and Step 5 Preview in 8 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in reasoning, where Step 5 Preview leads 40.0 to 8.6.
- The biggest single-benchmark swing is CritPt: 0.3% for Magistral Medium and 20.9% for Step 5 Preview.
- Step 5 Preview is cheaper at $1 / $2.70 per million input/output tokens, against $2 / $5 for Magistral Medium.
- Step 5 Preview accepts more context: 1.02M tokens versus 262K.
- Magistral Medium has downloadable open weights; the other is API-only.
Side by side
| Magistral Medium | Step 5 Preview | |
|---|---|---|
| Provider | Mistral AI | StepFun |
| Noometry Index | 35.2 | 47.9 |
| Released | 2025-03-17 | 2026-09-16 |
| Weights | Open | Proprietary |
| Context window | 262K | 1.02M |
| Max output | 16K | 66K |
| Input $ / M tokens | $2 | $1 |
| Output $ / M tokens | $5 | $2.70 |
| Results tracked | 22 | 18 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Step 5 Preview leads
Magistral Medium: 39.1 (#161), Step 5 Preview: 51.8 (#36)
| Benchmark | Magistral Medium | Step 5 Preview |
|---|---|---|
| SciCode | 39.2% | 58.9% |
| LMArena Coding | 1319 | 1480 |
| LMArena WebDev | — | 1564 |
Reasoning Step 5 Preview leads
Magistral Medium: 8.6 (#348), Step 5 Preview: 40.0 (#57)
| Benchmark | Magistral Medium | Step 5 Preview |
|---|---|---|
| CritPt | 0.3% | 20.9% |
| LMArena Hard Prompts | 1267 | 1465 |
| ARC-AGI-2 | 0% | — |
| Kagi LLM Benchmark | 16.2% | — |
| ARC-AGI-1 | 6.1% | — |
Math Step 5 Preview leads
Magistral Medium: 35.1 (#189), Step 5 Preview: 46.1 (#72)
| Benchmark | Magistral Medium | Step 5 Preview |
|---|---|---|
| LMArena Math | 1250 | 1470 |
| ProofBench | — | 42% |
Knowledge Step 5 Preview leads
Magistral Medium: 33.5 (#202), Step 5 Preview: 41.2 (#112)
| Benchmark | Magistral Medium | Step 5 Preview |
|---|---|---|
| LMArena Expert | 1223 | 1470 |
Multimodal Not comparable
Magistral Medium: —, Step 5 Preview: 41.0 (#41)
| Benchmark | Magistral Medium | Step 5 Preview |
|---|---|---|
| LMArena Vision | — | 1267 |
Multilingual Step 5 Preview leads
Magistral Medium: 39.6 (#224), Step 5 Preview: 53.8 (#54)
| Benchmark | Magistral Medium | Step 5 Preview |
|---|---|---|
| LMArena Non-English | 1232 | 1432 |
| LMArena Chinese | 1227 | 1519 |
| LMArena Russian | 1224 | 1436 |
| LMArena Spanish | 1271 | 1439 |
| LMArena French | 1267 | — |
| LMArena German | 1248 | — |
| LMArena Japanese | 1175 | — |
| LMArena Korean | 1125 | — |
Instruction Following Step 5 Preview leads
Magistral Medium: 66.0 (#211), Step 5 Preview: 76.0 (#50)
| Benchmark | Magistral Medium | Step 5 Preview |
|---|---|---|
| LMArena Instruction Following | 1254 | 1444 |
Long Context Step 5 Preview leads
Magistral Medium: 39.3 (#183), Step 5 Preview: 44.5 (#64)
| Benchmark | Magistral Medium | Step 5 Preview |
|---|---|---|
| LMArena Longer Query | 1295 | 1455 |
Writing & Preference Step 5 Preview leads
Magistral Medium: 46.3 (#219), Step 5 Preview: 62.8 (#71)
| Benchmark | Magistral Medium | Step 5 Preview |
|---|---|---|
| LMArena Text | 1255 | 1442 |
| LMArena Creative Writing | 1245 | 1410 |
| LMArena Multi-Turn | 1275 | 1447 |
Frequently asked questions
Is Magistral Medium better than Step 5 Preview?
Step 5 Preview is the stronger model overall, scoring 47.9 to 35.2 on the Noometry Index.
Which is cheaper, Magistral Medium or Step 5 Preview?
Step 5 Preview is cheaper. It lists at $1 per million input tokens and $2.70 per million output tokens; Magistral Medium lists at $2 and $5.
Is Magistral Medium or Step 5 Preview better for coding?
Step 5 Preview scores higher on coding benchmarks: 51.8 versus 39.1 in the Noometry coding category.
Which has the bigger context window?
Step 5 Preview does, with 1.02M tokens against 262K.
How many benchmarks do Magistral Medium and Step 5 Preview share?
15 benchmarks have published results for both models. Magistral Medium has 22 scored results on Noometry and Step 5 Preview has 18.