# Gemma 3 4B vs gpt-oss-20b

> gpt-oss-20b is the stronger model overall, scoring 32.1 to 27.9 on the Noometry Index.

- Canonical page: https://noometry.com/compare/gemma-3-4b-vs-gpt-oss-20b
- Last updated: 2026-10-10
- Shared benchmarks: 20

## Summary

- They share 20 benchmarks with published results for both. Gemma 3 4B scores higher in 5 categories and gpt-oss-20b in 4 categories; 7 gaps are clear of the uncertainty.
- The widest gap is in knowledge, where gpt-oss-20b leads 34.6 to 11.8.
- The biggest single-benchmark swing is OTIS Mock AIME 2024-2025: 7.5% for Gemma 3 4B and 65.3% for gpt-oss-20b.
- gpt-oss-20b is cheaper at $0.018 / $0.09 per million input/output tokens, against $0.04 / $0.08 for Gemma 3 4B.

## Snapshot

| | Gemma 3 4B | gpt-oss-20b |
|---|---|---|
| Provider | Google | OpenAI |
| Noometry Index | 27.9 | 32.1 |
| Rank | 320 | 249 |
| Context | 131K | 131K |
| Input $/M | $0.04 | $0.018 |
| Output $/M | $0.08 | $0.09 |
| Weights | Open | Open |

## Coding

- Gemma 3 4B: 35.3 (#214)
- gpt-oss-20b: 37.1 (#185)

| Benchmark | Gemma 3 4B | gpt-oss-20b |
|---|---|---|
| LMArena Coding | 1230 | 1307 |
| SciCode | — | 34.4% |
| WeirdML | — | 40.9% |
| ALE-Bench | — | 566.05 |

## Agentic & Tool Use

- Gemma 3 4B: 20.7 (#134)
- gpt-oss-20b: 9.2 (#146)

| Benchmark | Gemma 3 4B | gpt-oss-20b |
|---|---|---|
| Terminal-Bench | — | 3.4% |
| Berkeley Function Calling Leaderboard | 19.6% | — |

## Reasoning

- Gemma 3 4B: 12.7 (#330)
- gpt-oss-20b: 18.8 (#251)

| Benchmark | Gemma 3 4B | gpt-oss-20b |
|---|---|---|
| Kagi LLM Benchmark | 25.2% | 53.2% |
| Chess Puzzles | 0% | 4% |
| LMArena Hard Prompts | 1253 | 1274 |
| DTBench | 50.9% | 68% |
| LMCA | 2.8% | 14.5% |
| Epoch Capabilities Index | 116.02 | 137.81 |
| CritPt | — | 1.4% |

## Math

- Gemma 3 4B: 15.8 (#286)
- gpt-oss-20b: 37.6 (#93)

| Benchmark | Gemma 3 4B | gpt-oss-20b |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 7.5% | 65.3% |
| LMArena Math | 1239 | 1317 |
| Omni-MATH | — | 56.5% |

## Knowledge

- Gemma 3 4B: 11.8 (#294)
- gpt-oss-20b: 34.6 (#186)

| Benchmark | Gemma 3 4B | gpt-oss-20b |
|---|---|---|
| GPQA Diamond | 23.2% | 60.8% |
| LMArena Expert | 1223 | 1258 |
| MMLU-Pro | — | 74% |
| Vectara Hallucination Rate | 6.4% | — |
| GPQA (HELM) | — | 59.4% |

## Multilingual

- Gemma 3 4B: 42.7 (#189)
- gpt-oss-20b: 42.4 (#192)

| Benchmark | Gemma 3 4B | gpt-oss-20b |
|---|---|---|
| LMArena Non-English | 1273 | 1269 |
| LMArena German | 1281 | 1249 |
| LMArena Russian | 1294 | 1279 |
| LMArena Chinese | — | 1311 |
| LMArena Japanese | — | 1245 |
| LMArena Korean | — | 1239 |
| LMArena Spanish | — | 1267 |

## Instruction Following

- Gemma 3 4B: 65.4 (#219)
- gpt-oss-20b: 62.0 (#235)

| Benchmark | Gemma 3 4B | gpt-oss-20b |
|---|---|---|
| LMArena Instruction Following | 1239 | 1236 |
| IFEval | — | 73.2% |

## Long Context

- Gemma 3 4B: 38.6 (#189)
- gpt-oss-20b: 37.9 (#203)

| Benchmark | Gemma 3 4B | gpt-oss-20b |
|---|---|---|
| LMArena Longer Query | 1274 | 1251 |

## Writing & Preference

- Gemma 3 4B: 42.2 (#234)
- gpt-oss-20b: 35.8 (#260)

| Benchmark | Gemma 3 4B | gpt-oss-20b |
|---|---|---|
| LMArena Text | 1291 | 1288 |
| LMArena Creative Writing | 1270 | 1201 |
| EQ-Bench Creative Writing | 1068 | 666 |
| LMArena Multi-Turn | 1255 | 1268 |
| WildBench | — | 73.7% |

## FAQ

### Is Gemma 3 4B better than gpt-oss-20b?

gpt-oss-20b is the stronger model overall, scoring 32.1 to 27.9 on the Noometry Index.

### Which is cheaper, Gemma 3 4B or gpt-oss-20b?

gpt-oss-20b is cheaper. It lists at $0.018 per million input tokens and $0.09 per million output tokens; Gemma 3 4B lists at $0.04 and $0.08.

### Is Gemma 3 4B or gpt-oss-20b better for coding?

gpt-oss-20b scores higher on coding benchmarks: 37.1 versus 35.3 in the Noometry coding category.

### Which has the bigger context window?

Both accept 131K tokens.

### How many benchmarks do Gemma 3 4B and gpt-oss-20b share?

20 benchmarks have published results for both models. Gemma 3 4B has 22 scored results on Noometry and gpt-oss-20b has 34.