Model comparison
Gemma 4 31B IT vs Step 5 Preview
Step 5 Preview is the stronger model overall, scoring 47.9 to 43.5 on the Noometry Index. Gemma 4 31B IT costs 9.3× less per token, which makes it the better buy when Step 5 Preview's lead doesn't matter for your workload.
Last verified . 17 shared benchmarks.
Google
StepFunSummary
- They share 17 benchmarks with published results for both. Gemma 4 31B IT scores higher in 1 category and Step 5 Preview in 8 categories; 5 gaps are clear of the uncertainty.
- The widest gap is in reasoning, where Step 5 Preview leads 40.0 to 27.2.
- The biggest single-benchmark swing is CritPt: 1.4% for Gemma 4 31B IT and 20.9% for Step 5 Preview.
- Gemma 4 31B IT is cheaper at $0.09 / $0.34 per million input/output tokens, against $1 / $2.70 for Step 5 Preview.
- Step 5 Preview accepts more context: 1.02M tokens versus 262K.
- Gemma 4 31B IT has downloadable open weights; the other is API-only.
Side by side
| Gemma 4 31B IT | Step 5 Preview | |
|---|---|---|
| Provider | StepFun | |
| Noometry Index | 43.5 | 47.9 |
| Released | 2026-04-02 | 2026-09-16 |
| Weights | Open | Proprietary |
| Context window | 262K | 1.02M |
| Max output | 33K | 66K |
| Input $ / M tokens | $0.09 | $1 |
| Output $ / M tokens | $0.34 | $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
Gemma 4 31B IT: 42.3 (#108), Step 5 Preview: 51.8 (#36)
| Benchmark | Gemma 4 31B IT | Step 5 Preview |
|---|---|---|
| LMArena WebDev | 1366 | 1564 |
| SciCode | 43.4% | 58.9% |
| LMArena Coding | 1459 | 1480 |
| WeirdML | 52.3% | — |
| ALE-Bench | 925.5 | — |
Reasoning Step 5 Preview leads
Gemma 4 31B IT: 27.2 (#122), Step 5 Preview: 40.0 (#57)
| Benchmark | Gemma 4 31B IT | Step 5 Preview |
|---|---|---|
| CritPt | 1.4% | 20.9% |
| LMArena Hard Prompts | 1448 | 1465 |
| Kagi LLM Benchmark | 63.5% | — |
| NYT Connections (extended) | 70.6% | — |
| Chess Puzzles | 5% | — |
| Thematic Generalization | 53% | — |
| DTBench | 82.7% | — |
| LMCA | 39.3% | — |
| Surface Evolver Bench | 30.6% | — |
| Epoch Capabilities Index | 142.74 | — |
Math Step 5 Preview leads
Gemma 4 31B IT: 43.2 (#81), Step 5 Preview: 46.1 (#72)
| Benchmark | Gemma 4 31B IT | Step 5 Preview |
|---|---|---|
| LMArena Math | 1465 | 1470 |
| OTIS Mock AIME 2024-2025 | 73.3% | — |
| ProofBench | — | 42% |
Knowledge Step 5 Preview leads
Gemma 4 31B IT: 37.9 (#151), Step 5 Preview: 41.2 (#112)
| Benchmark | Gemma 4 31B IT | Step 5 Preview |
|---|---|---|
| LMArena Expert | 1465 | 1470 |
| GPQA Diamond | 75.8% | — |
| SimpleQA Verified | 10.4% | — |
| Vectara Hallucination Rate | 7.4% | — |
Multimodal Too close to call
Gemma 4 31B IT: 41.6 (#34), Step 5 Preview: 41.0 (#41)
| Benchmark | Gemma 4 31B IT | Step 5 Preview |
|---|---|---|
| LMArena Vision | 1277 | 1267 |
| LMArena Document | 1425 | — |
Multilingual Too close to call
Gemma 4 31B IT: 53.8 (#57), Step 5 Preview: 53.8 (#54)
| Benchmark | Gemma 4 31B IT | Step 5 Preview |
|---|---|---|
| LMArena Non-English | 1431 | 1432 |
| LMArena Chinese | 1476 | 1519 |
| LMArena Russian | 1460 | 1436 |
| LMArena Spanish | 1444 | 1439 |
| LMArena French | 1435 | — |
Instruction Following Too close to call
Gemma 4 31B IT: 75.5 (#61), Step 5 Preview: 76.0 (#50)
| Benchmark | Gemma 4 31B IT | Step 5 Preview |
|---|---|---|
| LMArena Instruction Following | 1433 | 1444 |
Long Context Too close to call
Gemma 4 31B IT: 44.2 (#71), Step 5 Preview: 44.5 (#64)
| Benchmark | Gemma 4 31B IT | Step 5 Preview |
|---|---|---|
| LMArena Longer Query | 1446 | 1455 |
Writing & Preference Step 5 Preview leads
Gemma 4 31B IT: 60.5 (#96), Step 5 Preview: 62.8 (#71)
| Benchmark | Gemma 4 31B IT | Step 5 Preview |
|---|---|---|
| LMArena Text | 1443 | 1442 |
| LMArena Creative Writing | 1415 | 1410 |
| LMArena Multi-Turn | 1452 | 1447 |
| EQ-Bench Creative Writing | 1368 | — |
| EQ-Bench 4 | 1120 | — |
Frequently asked questions
Is Gemma 4 31B IT better than Step 5 Preview?
Step 5 Preview is the stronger model overall, scoring 47.9 to 43.5 on the Noometry Index. Gemma 4 31B IT costs 9.3× less per token, which makes it the better buy when Step 5 Preview's lead doesn't matter for your workload.
Which is cheaper, Gemma 4 31B IT or Step 5 Preview?
Gemma 4 31B IT is cheaper. It lists at $0.09 per million input tokens and $0.34 per million output tokens; Step 5 Preview lists at $1 and $2.70.
Is Gemma 4 31B IT or Step 5 Preview better for coding?
Step 5 Preview scores higher on coding benchmarks: 51.8 versus 42.3 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 Gemma 4 31B IT and Step 5 Preview share?
17 benchmarks have published results for both models. Gemma 4 31B IT has 35 scored results on Noometry and Step 5 Preview has 18.