# Claude 3 Sonnet vs gpt-oss-20b

> gpt-oss-20b is the stronger model overall, scoring 32.1 to 28.8 on the Noometry Index.

- Canonical page: https://noometry.com/compare/claude-3-sonnet-vs-gpt-oss-20b
- Last updated: 2026-10-10
- Shared benchmarks: 21

## Summary

- They share 21 benchmarks with published results for both. Claude 3 Sonnet scores higher in 3 categories and gpt-oss-20b in 5 categories; 7 gaps are clear of the uncertainty.
- The widest gap is in math, where gpt-oss-20b leads 37.6 to 10.0.
- The biggest single-benchmark swing is OTIS Mock AIME 2024-2025: 2.5% for Claude 3 Sonnet and 65.3% for gpt-oss-20b.
- gpt-oss-20b has downloadable open weights; the other is API-only.

## Snapshot

| | Claude 3 Sonnet | gpt-oss-20b |
|---|---|---|
| Provider | Anthropic | OpenAI |
| Noometry Index | 28.8 | 32.1 |
| Rank | 311 | 249 |
| Context | — | 131K |
| Input $/M | — | $0.018 |
| Output $/M | — | $0.09 |
| Weights | Proprietary | Open |

## Coding

- Claude 3 Sonnet: 29.2 (#296)
- gpt-oss-20b: 37.1 (#185)

| Benchmark | Claude 3 Sonnet | gpt-oss-20b |
|---|---|---|
| WeirdML | 10.2% | 40.9% |
| LMArena Coding | 1223 | 1307 |
| SciCode | — | 34.4% |
| BigCodeBench Instruct | 42.7% | — |
| BigCodeBench Complete | 53.8% | — |
| ALE-Bench | — | 566.05 |
| HumanEval+ | 64% | — |
| MBPP+ | 69.3% | — |

## Agentic & Tool Use

- Claude 3 Sonnet: —
- gpt-oss-20b: 9.2 (#146)

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

## Reasoning

- Claude 3 Sonnet: 19.7 (#232)
- gpt-oss-20b: 18.8 (#251)

| Benchmark | Claude 3 Sonnet | gpt-oss-20b |
|---|---|---|
| LMArena Hard Prompts | 1197 | 1274 |
| DTBench | 53.6% | 68% |
| Epoch Capabilities Index | 120.69 | 137.81 |
| Kagi LLM Benchmark | — | 53.2% |
| CritPt | — | 1.4% |
| Chess Puzzles | — | 4% |
| LMCA | — | 14.5% |
| WinoGrande | 75.1% | — |

## Math

- Claude 3 Sonnet: 10.0 (#305)
- gpt-oss-20b: 37.6 (#93)

| Benchmark | Claude 3 Sonnet | gpt-oss-20b |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 2.5% | 65.3% |
| LMArena Math | 1213 | 1317 |
| Omni-MATH | — | 56.5% |
| MATH Level 5 | 18.2% | — |

## Knowledge

- Claude 3 Sonnet: 21.1 (#271)
- gpt-oss-20b: 34.6 (#186)

| Benchmark | Claude 3 Sonnet | gpt-oss-20b |
|---|---|---|
| GPQA Diamond | 40.6% | 60.8% |
| LMArena Expert | 1173 | 1258 |
| MMLU-Pro | — | 74% |
| GPQA (HELM) | — | 59.4% |
| MMLU | 75.9% | — |

## Multimodal

- Claude 3 Sonnet: 24.7 (#120)
- gpt-oss-20b: —

| Benchmark | Claude 3 Sonnet | gpt-oss-20b |
|---|---|---|
| LMArena Vision | 984 | — |

## Multilingual

- Claude 3 Sonnet: 38.0 (#229)
- gpt-oss-20b: 42.4 (#192)

| Benchmark | Claude 3 Sonnet | gpt-oss-20b |
|---|---|---|
| LMArena Non-English | 1205 | 1269 |
| LMArena Chinese | 1188 | 1311 |
| LMArena German | 1204 | 1249 |
| LMArena Japanese | 1131 | 1245 |
| LMArena Korean | 1128 | 1239 |
| LMArena Russian | 1227 | 1279 |
| LMArena Spanish | 1204 | 1267 |
| LMArena French | 1228 | — |

## Instruction Following

- Claude 3 Sonnet: 63.1 (#230)
- gpt-oss-20b: 62.0 (#235)

| Benchmark | Claude 3 Sonnet | gpt-oss-20b |
|---|---|---|
| LMArena Instruction Following | 1200 | 1236 |
| IFEval | — | 73.2% |

## Long Context

- Claude 3 Sonnet: 36.7 (#223)
- gpt-oss-20b: 37.9 (#203)

| Benchmark | Claude 3 Sonnet | gpt-oss-20b |
|---|---|---|
| LMArena Longer Query | 1211 | 1251 |

## Writing & Preference

- Claude 3 Sonnet: 42.3 (#233)
- gpt-oss-20b: 35.8 (#260)

| Benchmark | Claude 3 Sonnet | gpt-oss-20b |
|---|---|---|
| LMArena Text | 1218 | 1288 |
| LMArena Creative Writing | 1186 | 1201 |
| LMArena Multi-Turn | 1227 | 1268 |
| EQ-Bench Creative Writing | — | 666 |
| WildBench | — | 73.7% |

## FAQ

### Is Claude 3 Sonnet better than gpt-oss-20b?

gpt-oss-20b is the stronger model overall, scoring 32.1 to 28.8 on the Noometry Index.

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

gpt-oss-20b scores higher on coding benchmarks: 37.1 versus 29.2 in the Noometry coding category.

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

21 benchmarks have published results for both models. Claude 3 Sonnet has 30 scored results on Noometry and gpt-oss-20b has 34.