# gpt-oss-20b vs Step 3

> Step 3 is the stronger model overall, scoring 39.9 to 32.1 on the Noometry Index.

- Canonical page: https://noometry.com/compare/gpt-oss-20b-vs-step-3
- Last updated: 2026-10-10
- Shared benchmarks: 16

## Summary

- They share 16 benchmarks with published results for both. gpt-oss-20b scores higher in 1 category and Step 3 in 7 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in writing & preference, where Step 3 leads 54.5 to 35.8.
- The biggest single-benchmark swing is Kagi LLM Benchmark: 53.2% for gpt-oss-20b and 62.3% for Step 3.

## Snapshot

| | gpt-oss-20b | Step 3 |
|---|---|---|
| Provider | OpenAI | StepFun |
| Noometry Index | 32.1 | 39.9 |
| Rank | 249 | 145 |
| Context | 131K | — |
| Input $/M | $0.018 | — |
| Output $/M | $0.09 | — |
| Weights | Open | Open |

## Coding

- gpt-oss-20b: 37.1 (#185)
- Step 3: 39.5 (#144)

| Benchmark | gpt-oss-20b | Step 3 |
|---|---|---|
| LMArena Coding | 1307 | 1367 |
| SciCode | 34.4% | — |
| WeirdML | 40.9% | — |
| ALE-Bench | 566.05 | — |

## Agentic & Tool Use

- gpt-oss-20b: 9.2 (#146)
- Step 3: —

| Benchmark | gpt-oss-20b | Step 3 |
|---|---|---|
| Terminal-Bench | 3.4% | — |

## Reasoning

- gpt-oss-20b: 18.8 (#251)
- Step 3: 27.6 (#102)

| Benchmark | gpt-oss-20b | Step 3 |
|---|---|---|
| Kagi LLM Benchmark | 53.2% | 62.3% |
| LMArena Hard Prompts | 1274 | 1355 |
| CritPt | 1.4% | — |
| Chess Puzzles | 4% | — |
| DTBench | 68% | — |
| LMCA | 14.5% | — |
| Epoch Capabilities Index | 137.81 | — |

## Math

- gpt-oss-20b: 37.6 (#93)
- Step 3: 34.9 (#148)

| Benchmark | gpt-oss-20b | Step 3 |
|---|---|---|
| LMArena Math | 1317 | 1366 |
| OTIS Mock AIME 2024-2025 | 65.3% | — |
| Omni-MATH | 56.5% | — |

## Knowledge

- gpt-oss-20b: 34.6 (#186)
- Step 3: 36.2 (#161)

| Benchmark | gpt-oss-20b | Step 3 |
|---|---|---|
| LMArena Expert | 1258 | 1333 |
| GPQA Diamond | 60.8% | — |
| MMLU-Pro | 74% | — |
| GPQA (HELM) | 59.4% | — |

## Multimodal

- gpt-oss-20b: —
- Step 3: 34.9 (#81)

| Benchmark | gpt-oss-20b | Step 3 |
|---|---|---|
| LMArena Vision | — | 1176 |

## Multilingual

- gpt-oss-20b: 42.4 (#192)
- Step 3: 46.5 (#154)

| Benchmark | gpt-oss-20b | Step 3 |
|---|---|---|
| LMArena Non-English | 1269 | 1327 |
| LMArena Chinese | 1311 | 1396 |
| LMArena German | 1249 | 1371 |
| LMArena Korean | 1239 | 1271 |
| LMArena Russian | 1279 | 1331 |
| LMArena Spanish | 1267 | 1372 |
| LMArena Japanese | 1245 | — |

## Instruction Following

- gpt-oss-20b: 62.0 (#235)
- Step 3: 70.6 (#159)

| Benchmark | gpt-oss-20b | Step 3 |
|---|---|---|
| LMArena Instruction Following | 1236 | 1332 |
| IFEval | 73.2% | — |

## Long Context

- gpt-oss-20b: 37.9 (#203)
- Step 3: 40.3 (#152)

| Benchmark | gpt-oss-20b | Step 3 |
|---|---|---|
| LMArena Longer Query | 1251 | 1326 |

## Writing & Preference

- gpt-oss-20b: 35.8 (#260)
- Step 3: 54.5 (#146)

| Benchmark | gpt-oss-20b | Step 3 |
|---|---|---|
| LMArena Text | 1288 | 1351 |
| LMArena Creative Writing | 1201 | 1322 |
| LMArena Multi-Turn | 1268 | 1340 |
| EQ-Bench Creative Writing | 666 | — |
| WildBench | 73.7% | — |

## FAQ

### Is gpt-oss-20b better than Step 3?

Step 3 is the stronger model overall, scoring 39.9 to 32.1 on the Noometry Index.

### Is gpt-oss-20b or Step 3 better for coding?

Step 3 scores higher on coding benchmarks: 39.5 versus 37.1 in the Noometry coding category.

### How many benchmarks do gpt-oss-20b and Step 3 share?

16 benchmarks have published results for both models. gpt-oss-20b has 34 scored results on Noometry and Step 3 has 17.