Model comparison
Gemma 3 4B vs Step 5 Preview
Step 5 Preview is the stronger model overall, scoring 47.9 to 28.1 on the Noometry Index. Gemma 3 4B costs 29× less per token, which makes it the better buy when Step 5 Preview's lead doesn't matter for your workload.
Last verified . 11 shared benchmarks.
Google
StepFunSummary
- They share 11 benchmarks with published results for both. Gemma 3 4B scores higher in 0 categories and Step 5 Preview in 8 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in knowledge, where Step 5 Preview leads 41.2 to 11.8.
- Gemma 3 4B is cheaper at $0.04 / $0.08 per million input/output tokens, against $1 / $2.70 for Step 5 Preview.
- Step 5 Preview accepts more context: 1.02M tokens versus 131K.
- Gemma 3 4B has downloadable open weights; the other is API-only.
Side by side
| Gemma 3 4B | Step 5 Preview | |
|---|---|---|
| Provider | StepFun | |
| Noometry Index | 28.1 | 47.9 |
| Released | 2025-03-12 | 2026-09-16 |
| Weights | Open | Proprietary |
| Context window | 131K | 1.02M |
| Max output | 4K | 66K |
| Input $ / M tokens | $0.04 | $1 |
| Output $ / M tokens | $0.08 | $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
Gemma 3 4B: 35.9 (#215), Step 5 Preview: 51.8 (#36)
| Benchmark | Gemma 3 4B | Step 5 Preview |
|---|---|---|
| LMArena Coding | 1230 | 1480 |
| LMArena WebDev | — | 1564 |
| SciCode | — | 58.9% |
Agentic & Tool Use Not comparable
Gemma 3 4B: 20.9 (#142), Step 5 Preview: —
| Benchmark | Gemma 3 4B | Step 5 Preview |
|---|---|---|
| Berkeley Function Calling Leaderboard | 19.6% | — |
Reasoning Step 5 Preview leads
Gemma 3 4B: 13.2 (#335), Step 5 Preview: 40.0 (#57)
| Benchmark | Gemma 3 4B | Step 5 Preview |
|---|---|---|
| LMArena Hard Prompts | 1253 | 1465 |
| Kagi LLM Benchmark | 25.2% | — |
| CritPt | — | 20.9% |
| Chess Puzzles | 0% | — |
| DTBench | 50.9% | — |
| LMCA | 2.8% | — |
| Epoch Capabilities Index | 116.02 | — |
Math Step 5 Preview leads
Gemma 3 4B: 16.8 (#292), Step 5 Preview: 46.1 (#72)
| Benchmark | Gemma 3 4B | Step 5 Preview |
|---|---|---|
| LMArena Math | 1239 | 1470 |
| OTIS Mock AIME 2024-2025 | 7.5% | — |
| ProofBench | — | 42% |
Knowledge Step 5 Preview leads
Gemma 3 4B: 11.8 (#299), Step 5 Preview: 41.2 (#112)
| Benchmark | Gemma 3 4B | Step 5 Preview |
|---|---|---|
| LMArena Expert | 1223 | 1470 |
| GPQA Diamond | 23.2% | — |
| Vectara Hallucination Rate | 6.4% | — |
Multimodal Not comparable
Gemma 3 4B: —, Step 5 Preview: 41.0 (#41)
| Benchmark | Gemma 3 4B | Step 5 Preview |
|---|---|---|
| LMArena Vision | — | 1267 |
Multilingual Step 5 Preview leads
Gemma 3 4B: 42.5 (#194), Step 5 Preview: 53.8 (#54)
| Benchmark | Gemma 3 4B | Step 5 Preview |
|---|---|---|
| LMArena Non-English | 1273 | 1432 |
| LMArena Russian | 1294 | 1436 |
| LMArena Chinese | — | 1519 |
| LMArena German | 1281 | — |
| LMArena Spanish | — | 1439 |
Instruction Following Step 5 Preview leads
Gemma 3 4B: 65.2 (#225), Step 5 Preview: 76.0 (#50)
| Benchmark | Gemma 3 4B | Step 5 Preview |
|---|---|---|
| LMArena Instruction Following | 1239 | 1444 |
Long Context Step 5 Preview leads
Gemma 3 4B: 38.7 (#194), Step 5 Preview: 44.5 (#64)
| Benchmark | Gemma 3 4B | Step 5 Preview |
|---|---|---|
| LMArena Longer Query | 1273 | 1455 |
Writing & Preference Step 5 Preview leads
Gemma 3 4B: 42.0 (#239), Step 5 Preview: 62.8 (#71)
| Benchmark | Gemma 3 4B | Step 5 Preview |
|---|---|---|
| LMArena Text | 1291 | 1442 |
| LMArena Creative Writing | 1271 | 1410 |
| LMArena Multi-Turn | 1255 | 1447 |
| EQ-Bench Creative Writing | 1068 | — |
Frequently asked questions
Is Gemma 3 4B better than Step 5 Preview?
Step 5 Preview is the stronger model overall, scoring 47.9 to 28.1 on the Noometry Index. Gemma 3 4B costs 29× 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 3 4B or Step 5 Preview?
Gemma 3 4B is cheaper. It lists at $0.04 per million input tokens and $0.08 per million output tokens; Step 5 Preview lists at $1 and $2.70.
Is Gemma 3 4B or Step 5 Preview better for coding?
Step 5 Preview scores higher on coding benchmarks: 51.8 versus 35.9 in the Noometry coding category.
Which has the bigger context window?
Step 5 Preview does, with 1.02M tokens against 131K.
How many benchmarks do Gemma 3 4B and Step 5 Preview share?
11 benchmarks have published results for both models. Gemma 3 4B has 22 scored results on Noometry and Step 5 Preview has 18.