# gpt-oss-20b vs Grok 4

> Grok 4 is the stronger model overall, scoring 48.1 to 32.5 on the Noometry Index.

- Canonical page: https://noometry.com/compare/gpt-oss-20b-vs-grok-4
- Last updated: 2026-10-11
- Shared benchmarks: 28

## Summary

- They share 28 benchmarks with published results for both. gpt-oss-20b scores higher in 0 categories and Grok 4 in 9 categories; 9 gaps are clear of the uncertainty.
- The widest gap is in long context, where Grok 4 leads 63.1 to 37.9.
- The biggest single-benchmark swing is GPQA Diamond: 60.8% for gpt-oss-20b and 87% for Grok 4.
- gpt-oss-20b has downloadable open weights; the other is API-only.

## Snapshot

| | gpt-oss-20b | Grok 4 |
|---|---|---|
| Provider | OpenAI | xAI |
| Noometry Index | 32.5 | 48.1 |
| Rank | 255 | 56 |
| Context | 131K | — |
| Input $/M | $0.018 | — |
| Output $/M | $0.09 | — |
| Weights | Open | Proprietary |

## Coding

- gpt-oss-20b: 37.6 (#192)
- Grok 4: 50.3 (#46)

| Benchmark | gpt-oss-20b | Grok 4 |
|---|---|---|
| WeirdML | 40.9% | 45.7% |
| LMArena Coding | 1306 | 1408 |
| Aider Polyglot | — | 79.6% |
| SciCode | 34.4% | — |
| ALE-Bench | 566.05 | — |

## Agentic & Tool Use

- gpt-oss-20b: 9.3 (#154)
- Grok 4: 32.3 (#68)

| Benchmark | gpt-oss-20b | Grok 4 |
|---|---|---|
| Terminal-Bench | 3.4% | 27.2% |
| Berkeley Function Calling Leaderboard | — | 63% |
| GDPval | — | 21.1% |
| Cybench | — | 43% |
| DeepResearch Bench | — | 47.3% |
| BALROG | — | 43.6% |
| LMArena Search | — | 1142 |
| METR Time Horizons | — | 66.6% |

## Reasoning

- gpt-oss-20b: 19.3 (#261)
- Grok 4: 36.7 (#65)

| Benchmark | gpt-oss-20b | Grok 4 |
|---|---|---|
| Kagi LLM Benchmark | 53.2% | 73.6% |
| Chess Puzzles | 4% | 28% |
| LMArena Hard Prompts | 1274 | 1409 |
| Epoch Capabilities Index | 137.82 | 146.44 |
| ARC-AGI-2 | — | 16% |
| SimpleBench | — | 60.5% |
| ARC-AGI-1 | — | 66.7% |
| CritPt | 1.4% | — |
| DTBench | 68% | — |
| LMCA | 14.5% | — |
| ForecastBench | — | 60.9 |

## Math

- gpt-oss-20b: 39.4 (#103)
- Grok 4: 48.4 (#64)

| Benchmark | gpt-oss-20b | Grok 4 |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 65.3% | 84% |
| Omni-MATH | 56.5% | 60.3% |
| LMArena Math | 1317 | 1422 |
| FrontierMath (Feb 2025 set) | — | 19.7% |
| FrontierMath Tier 4 (v1) | — | 2.1% |

## Knowledge

- gpt-oss-20b: 34.6 (#195)
- Grok 4: 53.8 (#55)

| Benchmark | gpt-oss-20b | Grok 4 |
|---|---|---|
| GPQA Diamond | 60.8% | 87% |
| MMLU-Pro | 74% | 85.1% |
| GPQA (HELM) | 59.4% | 72.7% |
| LMArena Expert | 1258 | 1415 |
| Confabulations | — | 12.4% |

## Multimodal

- gpt-oss-20b: —
- Grok 4: 33.7 (#94)

| Benchmark | gpt-oss-20b | Grok 4 |
|---|---|---|
| LMArena Vision | — | 1210 |
| GeoBench | — | 45% |

## Multilingual

- gpt-oss-20b: 42.2 (#197)
- Grok 4: 51.8 (#103)

| Benchmark | gpt-oss-20b | Grok 4 |
|---|---|---|
| LMArena Non-English | 1268 | 1403 |
| LMArena Chinese | 1314 | 1427 |
| LMArena German | 1255 | 1429 |
| LMArena Japanese | 1244 | 1394 |
| LMArena Korean | 1236 | 1377 |
| LMArena Russian | 1278 | 1410 |
| LMArena Spanish | 1267 | 1420 |
| LMArena French | — | 1418 |

## Instruction Following

- gpt-oss-20b: 61.8 (#240)
- Grok 4: 79.2 (#5)

| Benchmark | gpt-oss-20b | Grok 4 |
|---|---|---|
| IFEval | 73.2% | 94.9% |
| LMArena Instruction Following | 1236 | 1387 |

## Long Context

- gpt-oss-20b: 37.9 (#209)
- Grok 4: 63.1 (#4)

| Benchmark | gpt-oss-20b | Grok 4 |
|---|---|---|
| LMArena Longer Query | 1250 | 1409 |
| Fiction.LiveBench | — | 94.4% |

## Writing & Preference

- gpt-oss-20b: 35.5 (#265)
- Grok 4: 58.5 (#116)

| Benchmark | gpt-oss-20b | Grok 4 |
|---|---|---|
| LMArena Text | 1287 | 1411 |
| LMArena Creative Writing | 1201 | 1397 |
| WildBench | 73.7% | 79.7% |
| LMArena Multi-Turn | 1268 | 1416 |
| Short-Story Creative Writing | — | 76.9% |
| EQ-Bench Creative Writing | 666 | — |

## FAQ

### Is gpt-oss-20b better than Grok 4?

Grok 4 is the stronger model overall, scoring 48.1 to 32.5 on the Noometry Index.

### Is gpt-oss-20b or Grok 4 better for coding?

Grok 4 scores higher on coding benchmarks: 50.3 versus 37.6 in the Noometry coding category.

### How many benchmarks do gpt-oss-20b and Grok 4 share?

28 benchmarks have published results for both models. gpt-oss-20b has 34 scored results on Noometry and Grok 4 has 48.