Model comparison
Grok-3 mini vs Step 5 Preview
Step 5 Preview is the stronger model overall, scoring 47.9 to 41.2 on the Noometry Index.
Last verified . 13 shared benchmarks.
StepFunSummary
- They share 13 benchmarks with published results for both. Grok-3 mini scores higher in 2 categories and Step 5 Preview in 6 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in reasoning, where Step 5 Preview leads 40.0 to 13.6.
Side by side
| Grok-3 mini | Step 5 Preview | |
|---|---|---|
| Provider | xAI | StepFun |
| Noometry Index | 41.2 | 47.9 |
| Released | 2025-04-09 | 2026-09-16 |
| Weights | Proprietary | Proprietary |
| Context window | — | 1.02M |
| Max output | — | 66K |
| Input $ / M tokens | — | $1 |
| Output $ / M tokens | — | $2.70 |
| Results tracked | 35 | 18 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Step 5 Preview leads
Grok-3 mini: 40.8 (#131), Step 5 Preview: 51.8 (#36)
| Benchmark | Grok-3 mini | Step 5 Preview |
|---|---|---|
| LMArena Coding | 1379 | 1480 |
| Aider Polyglot | 49.3% | — |
| LMArena WebDev | — | 1564 |
| SciCode | — | 58.9% |
| WeirdML | 42.6% | — |
Reasoning Step 5 Preview leads
Grok-3 mini: 13.6 (#334), Step 5 Preview: 40.0 (#57)
| Benchmark | Grok-3 mini | Step 5 Preview |
|---|---|---|
| LMArena Hard Prompts | 1375 | 1465 |
| ARC-AGI-2 | 0.4% | — |
| Kagi LLM Benchmark | 61.3% | — |
| ARC-AGI-1 | 16.5% | — |
| CritPt | — | 20.9% |
| Epoch Capabilities Index | 140.35 | — |
Math Step 5 Preview leads
Grok-3 mini: 42.1 (#85), Step 5 Preview: 46.1 (#72)
| Benchmark | Grok-3 mini | Step 5 Preview |
|---|---|---|
| LMArena Math | 1386 | 1470 |
| OTIS Mock AIME 2024-2025 | 77.8% | — |
| ProofBench | — | 42% |
| Omni-MATH | 31.8% | — |
| MATH Level 5 | 90.9% | — |
| FrontierMath (Feb 2025 set) | 5.9% | — |
Knowledge Grok-3 mini leads
Grok-3 mini: 46.4 (#81), Step 5 Preview: 41.2 (#112)
| Benchmark | Grok-3 mini | Step 5 Preview |
|---|---|---|
| LMArena Expert | 1395 | 1470 |
| GPQA Diamond | 76.3% | — |
| MMLU-Pro | 79.9% | — |
| Confabulations | 10.8% | — |
| GPQA (HELM) | 67.5% | — |
Multimodal Not comparable
Grok-3 mini: —, Step 5 Preview: 41.0 (#41)
| Benchmark | Grok-3 mini | Step 5 Preview |
|---|---|---|
| LMArena Vision | — | 1267 |
Multilingual Step 5 Preview leads
Grok-3 mini: 48.1 (#145), Step 5 Preview: 53.8 (#54)
| Benchmark | Grok-3 mini | Step 5 Preview |
|---|---|---|
| LMArena Non-English | 1352 | 1432 |
| LMArena Chinese | 1387 | 1519 |
| LMArena Russian | 1353 | 1436 |
| LMArena Spanish | 1381 | 1439 |
| LMArena French | 1357 | — |
| LMArena German | 1349 | — |
| LMArena Japanese | 1342 | — |
| LMArena Korean | 1335 | — |
Instruction Following Grok-3 mini leads
Grok-3 mini: 78.5 (#9), Step 5 Preview: 76.0 (#50)
| Benchmark | Grok-3 mini | Step 5 Preview |
|---|---|---|
| LMArena Instruction Following | 1357 | 1444 |
| IFEval | 95.1% | — |
Long Context Step 5 Preview leads
Grok-3 mini: 41.0 (#147), Step 5 Preview: 44.5 (#64)
| Benchmark | Grok-3 mini | Step 5 Preview |
|---|---|---|
| LMArena Longer Query | 1372 | 1455 |
| Fiction.LiveBench | 66.7% | — |
Writing & Preference Step 5 Preview leads
Grok-3 mini: 52.5 (#169), Step 5 Preview: 62.8 (#71)
| Benchmark | Grok-3 mini | Step 5 Preview |
|---|---|---|
| LMArena Text | 1370 | 1442 |
| LMArena Creative Writing | 1342 | 1410 |
| LMArena Multi-Turn | 1355 | 1447 |
| Short-Story Creative Writing | 73.5% | — |
| WildBench | 65.1% | — |
Frequently asked questions
Is Grok-3 mini better than Step 5 Preview?
Step 5 Preview is the stronger model overall, scoring 47.9 to 41.2 on the Noometry Index.
Is Grok-3 mini or Step 5 Preview better for coding?
Step 5 Preview scores higher on coding benchmarks: 51.8 versus 40.8 in the Noometry coding category.
How many benchmarks do Grok-3 mini and Step 5 Preview share?
13 benchmarks have published results for both models. Grok-3 mini has 35 scored results on Noometry and Step 5 Preview has 18.