# gpt-oss-120b vs Step 5 Preview

> Step 5 Preview is the stronger model overall, scoring 47.9 to 36.3 on the Noometry Index. gpt-oss-120b costs 20× less per token, which makes it the better buy when Step 5 Preview's lead doesn't matter for your workload.

- Canonical page: https://noometry.com/compare/gpt-oss-120b-vs-step-5-preview
- Last updated: 2026-10-10
- Shared benchmarks: 15

## Summary

- They share 15 benchmarks with published results for both. gpt-oss-120b 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 20.0.
- The biggest single-benchmark swing is SciCode: 36% for gpt-oss-120b and 58.9% for Step 5 Preview.
- gpt-oss-120b is cheaper at $0.037 / $0.17 per million input/output tokens, against $1 / $2.70 for Step 5 Preview.
- Step 5 Preview accepts more context: 1.02M tokens versus 131K.
- gpt-oss-120b has downloadable open weights; the other is API-only.

## Snapshot

| | gpt-oss-120b | Step 5 Preview |
|---|---|---|
| Provider | OpenAI | StepFun |
| Noometry Index | 36.3 | 47.9 |
| Rank | 217 | 58 |
| Context | 131K | 1.02M |
| Input $/M | $0.037 | $1 |
| Output $/M | $0.17 | $2.70 |
| Weights | Open | Proprietary |

## Coding

- gpt-oss-120b: 33.5 (#256)
- Step 5 Preview: 51.8 (#36)

| Benchmark | gpt-oss-120b | Step 5 Preview |
|---|---|---|
| SciCode | 36% | 58.9% |
| LMArena Coding | 1380 | 1480 |
| SWE-bench Verified (bash only) | 26% | — |
| Aider Polyglot | 41.8% | — |
| LMArena WebDev | — | 1564 |
| WeirdML | 48.2% | — |
| ALE-Bench | 575.62 | — |
| AlgoTune | 1.41 | — |

## Agentic & Tool Use

- gpt-oss-120b: 12.2 (#153)
- Step 5 Preview: —

| Benchmark | gpt-oss-120b | Step 5 Preview |
|---|---|---|
| Terminal-Bench | 18.7% | — |
| APEX-Agents | 4.4% | — |
| METR Time Horizons | 56.6% | — |
| Vending-Bench 2 | -21.53 | — |

## Reasoning

- gpt-oss-120b: 20.0 (#245)
- Step 5 Preview: 40.0 (#57)

| Benchmark | gpt-oss-120b | Step 5 Preview |
|---|---|---|
| CritPt | 1.1% | 20.9% |
| LMArena Hard Prompts | 1364 | 1465 |
| SimpleBench | 22.1% | — |
| Kagi LLM Benchmark | 58.6% | — |
| Chess Puzzles | 20% | — |
| Mystery Game Puzzles | 2% | — |
| DTBench | 76.3% | — |
| LMCA | 22.1% | — |
| Surface Evolver Bench | 25% | — |
| Epoch Capabilities Index | 139.93 | — |

## Math

- gpt-oss-120b: 52.5 (#50)
- Step 5 Preview: 46.1 (#72)

| Benchmark | gpt-oss-120b | Step 5 Preview |
|---|---|---|
| LMArena Math | 1389 | 1470 |
| OTIS Mock AIME 2024-2025 | 88.9% | — |
| ProofBench | — | 42% |
| Omni-MATH | 68.8% | — |

## Knowledge

- gpt-oss-120b: 42.4 (#96)
- Step 5 Preview: 41.2 (#112)

| Benchmark | gpt-oss-120b | Step 5 Preview |
|---|---|---|
| LMArena Expert | 1356 | 1470 |
| GPQA Diamond | 75.8% | — |
| MMLU-Pro | 79.5% | — |
| Confabulations | 15.7% | — |
| Vectara Hallucination Rate | 14.2% | — |
| GPQA (HELM) | 68.4% | — |

## Multimodal

- gpt-oss-120b: —
- Step 5 Preview: 41.0 (#41)

| Benchmark | gpt-oss-120b | Step 5 Preview |
|---|---|---|
| LMArena Vision | — | 1267 |

## Multilingual

- gpt-oss-120b: 48.0 (#147)
- Step 5 Preview: 53.8 (#54)

| Benchmark | gpt-oss-120b | Step 5 Preview |
|---|---|---|
| LMArena Non-English | 1351 | 1432 |
| LMArena Chinese | 1385 | 1519 |
| LMArena Russian | 1343 | 1436 |
| LMArena Spanish | 1389 | 1439 |
| LMArena French | 1369 | — |
| LMArena German | 1353 | — |
| LMArena Japanese | 1331 | — |
| LMArena Korean | 1282 | — |

## Instruction Following

- gpt-oss-120b: 69.3 (#173)
- Step 5 Preview: 76.0 (#50)

| Benchmark | gpt-oss-120b | Step 5 Preview |
|---|---|---|
| LMArena Instruction Following | 1318 | 1444 |
| IFEval | 83.6% | — |

## Long Context

- gpt-oss-120b: 31.4 (#278)
- Step 5 Preview: 44.5 (#64)

| Benchmark | gpt-oss-120b | Step 5 Preview |
|---|---|---|
| LMArena Longer Query | 1319 | 1455 |
| Fiction.LiveBench | 44.4% | — |

## Writing & Preference

- gpt-oss-120b: 46.5 (#217)
- Step 5 Preview: 62.8 (#71)

| Benchmark | gpt-oss-120b | Step 5 Preview |
|---|---|---|
| LMArena Text | 1365 | 1442 |
| LMArena Creative Writing | 1275 | 1410 |
| LMArena Multi-Turn | 1340 | 1447 |
| Short-Story Creative Writing | 77.1% | — |
| EQ-Bench Creative Writing | 961 | — |
| WildBench | 84.5% | — |

## FAQ

### Is gpt-oss-120b better than Step 5 Preview?

Step 5 Preview is the stronger model overall, scoring 47.9 to 36.3 on the Noometry Index. gpt-oss-120b costs 20× less per token, which makes it the better buy when Step 5 Preview's lead doesn't matter for your workload.

### Which is cheaper, gpt-oss-120b or Step 5 Preview?

gpt-oss-120b is cheaper. It lists at $0.037 per million input tokens and $0.17 per million output tokens; Step 5 Preview lists at $1 and $2.70.

### Is gpt-oss-120b or Step 5 Preview better for coding?

Step 5 Preview scores higher on coding benchmarks: 51.8 versus 33.5 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 gpt-oss-120b and Step 5 Preview share?

15 benchmarks have published results for both models. gpt-oss-120b has 48 scored results on Noometry and Step 5 Preview has 18.