# Gemini 3.8 Flash vs gpt-oss-20b

> Gemini 3.8 Flash is the stronger model overall, scoring 59.5 to 32.1 on the Noometry Index. gpt-oss-20b costs 42× less per token, which makes it the better buy when Gemini 3.8 Flash's lead doesn't matter for your workload.

- Canonical page: https://noometry.com/compare/gemini-3-8-flash-vs-gpt-oss-20b
- Last updated: 2026-10-10
- Shared benchmarks: 26

## Summary

- They share 26 benchmarks with published results for both. Gemini 3.8 Flash scores higher in 9 categories and gpt-oss-20b in 0 categories; 9 gaps are clear of the uncertainty.
- The widest gap is in reasoning, where Gemini 3.8 Flash leads 64.7 to 18.8.
- The biggest single-benchmark swing is Chess Puzzles: 61% for Gemini 3.8 Flash and 4% for gpt-oss-20b.
- gpt-oss-20b is cheaper at $0.018 / $0.09 per million input/output tokens, against $0.75 / $3.75 for Gemini 3.8 Flash.
- Gemini 3.8 Flash accepts more context: 1.05M tokens versus 131K.
- gpt-oss-20b has downloadable open weights; the other is API-only.

## Snapshot

| | Gemini 3.8 Flash | gpt-oss-20b |
|---|---|---|
| Provider | Google | OpenAI |
| Noometry Index | 59.5 | 32.1 |
| Rank | 13 | 249 |
| Context | 1.05M | 131K |
| Input $/M | $0.75 | $0.018 |
| Output $/M | $3.75 | $0.09 |
| Weights | Proprietary | Open |

## Coding

- Gemini 3.8 Flash: 56.4 (#17)
- gpt-oss-20b: 37.1 (#185)

| Benchmark | Gemini 3.8 Flash | gpt-oss-20b |
|---|---|---|
| SciCode | 56.6% | 34.4% |
| LMArena Coding | 1512 | 1307 |
| ALE-Bench | 1,270 | 566.05 |
| DeepSWE | 73.8% | — |
| FrontierCode | 41.2% | — |
| CursorBench | 39.6% | — |
| LMArena WebDev | 1581 | — |
| FrontierSWE | 19.6% | — |
| WeirdML | — | 40.9% |

## Agentic & Tool Use

- Gemini 3.8 Flash: 40.3 (#22)
- gpt-oss-20b: 9.2 (#146)

| Benchmark | Gemini 3.8 Flash | gpt-oss-20b |
|---|---|---|
| Terminal-Bench | — | 3.4% |
| APEX-Agents | 64.3% | — |
| GDP.pdf | 23.4% | — |
| Vending-Bench 2 | 5,094 | — |

## Reasoning

- Gemini 3.8 Flash: 64.7 (#13)
- gpt-oss-20b: 18.8 (#251)

| Benchmark | Gemini 3.8 Flash | gpt-oss-20b |
|---|---|---|
| CritPt | 18.3% | 1.4% |
| Chess Puzzles | 61% | 4% |
| LMArena Hard Prompts | 1506 | 1274 |
| DTBench | 95.7% | 68% |
| LMCA | 52.9% | 14.5% |
| Epoch Capabilities Index | 157.13 | 137.81 |
| Kagi LLM Benchmark | — | 53.2% |
| NYT Connections (extended) | 97.4% | — |
| Mystery Game Puzzles | 47% | — |
| Surface Evolver Bench | 76.9% | — |

## Math

- Gemini 3.8 Flash: 65.2 (#22)
- gpt-oss-20b: 37.6 (#93)

| Benchmark | Gemini 3.8 Flash | gpt-oss-20b |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 98.9% | 65.3% |
| LMArena Math | 1528 | 1317 |
| FrontierMath (Tiers 1-3) | 68.4% | — |
| FrontierMath Tier 4 | 22% | — |
| ProofBench | 48% | — |
| Omni-MATH | — | 56.5% |

## Knowledge

- Gemini 3.8 Flash: 74.9 (#2)
- gpt-oss-20b: 34.6 (#186)

| Benchmark | Gemini 3.8 Flash | gpt-oss-20b |
|---|---|---|
| GPQA Diamond | 95.4% | 60.8% |
| LMArena Expert | 1529 | 1258 |
| Humanity's Last Exam | 44.5% | — |
| SimpleQA Verified | 69.7% | — |
| MMLU-Pro | — | 74% |
| GPQA (HELM) | — | 59.4% |

## Multimodal

- Gemini 3.8 Flash: 40.9 (#33)
- gpt-oss-20b: —

| Benchmark | Gemini 3.8 Flash | gpt-oss-20b |
|---|---|---|
| LMArena Vision | 1312 | — |
| Blueprint-Bench 2 | 38.6% | — |
| Furniture Assembly | 31.7% | — |

## Multilingual

- Gemini 3.8 Flash: 57.9 (#5)
- gpt-oss-20b: 42.4 (#192)

| Benchmark | Gemini 3.8 Flash | gpt-oss-20b |
|---|---|---|
| LMArena Non-English | 1486 | 1269 |
| LMArena Chinese | 1547 | 1311 |
| LMArena German | 1470 | 1249 |
| LMArena Japanese | 1505 | 1245 |
| LMArena Korean | 1479 | 1239 |
| LMArena Russian | 1507 | 1279 |
| LMArena Spanish | 1492 | 1267 |
| LMArena French | 1502 | — |

## Instruction Following

- Gemini 3.8 Flash: 78.2 (#11)
- gpt-oss-20b: 62.0 (#235)

| Benchmark | Gemini 3.8 Flash | gpt-oss-20b |
|---|---|---|
| LMArena Instruction Following | 1489 | 1236 |
| IFEval | — | 73.2% |

## Long Context

- Gemini 3.8 Flash: 46.2 (#23)
- gpt-oss-20b: 37.9 (#203)

| Benchmark | Gemini 3.8 Flash | gpt-oss-20b |
|---|---|---|
| LMArena Longer Query | 1506 | 1251 |

## Writing & Preference

- Gemini 3.8 Flash: 72.1 (#16)
- gpt-oss-20b: 35.8 (#260)

| Benchmark | Gemini 3.8 Flash | gpt-oss-20b |
|---|---|---|
| LMArena Text | 1494 | 1288 |
| LMArena Creative Writing | 1485 | 1201 |
| EQ-Bench Creative Writing | 1748 | 666 |
| LMArena Multi-Turn | 1503 | 1268 |
| WildBench | — | 73.7% |

## FAQ

### Is Gemini 3.8 Flash better than gpt-oss-20b?

Gemini 3.8 Flash is the stronger model overall, scoring 59.5 to 32.1 on the Noometry Index. gpt-oss-20b costs 42× less per token, which makes it the better buy when Gemini 3.8 Flash's lead doesn't matter for your workload.

### Which is cheaper, Gemini 3.8 Flash 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; Gemini 3.8 Flash lists at $0.75 and $3.75.

### Is Gemini 3.8 Flash or gpt-oss-20b better for coding?

Gemini 3.8 Flash scores higher on coding benchmarks: 56.4 versus 37.1 in the Noometry coding category.

### Which has the bigger context window?

Gemini 3.8 Flash does, with 1.05M tokens against 131K.

### How many benchmarks do Gemini 3.8 Flash and gpt-oss-20b share?

26 benchmarks have published results for both models. Gemini 3.8 Flash has 46 scored results on Noometry and gpt-oss-20b has 34.