# gpt-oss-20b vs o3-pro

> o3-pro is the stronger model overall, scoring 42.9 to 32.5 on the Noometry Index. gpt-oss-20b costs 972× less per token, which makes it the better buy when o3-pro's lead doesn't matter for your workload.

- Canonical page: https://noometry.com/compare/gpt-oss-20b-vs-o3-pro
- Last updated: 2026-10-11
- Shared benchmarks: 5

## Summary

- They share 5 benchmarks with published results for both. gpt-oss-20b scores higher in 1 category and o3-pro in 4 categories; 5 gaps are clear of the uncertainty.
- The widest gap is in long context, where o3-pro leads 72.2 to 37.9.
- The biggest single-benchmark swing is LMCA: 14.5% for gpt-oss-20b and 38.5% for o3-pro.
- gpt-oss-20b is cheaper at $0.018 / $0.09 per million input/output tokens, against $20 / $80 for o3-pro.
- o3-pro accepts more context: 200K tokens versus 131K.
- gpt-oss-20b has downloadable open weights; the other is API-only.

## Snapshot

| | gpt-oss-20b | o3-pro |
|---|---|---|
| Provider | OpenAI | OpenAI |
| Noometry Index | 32.5 | 42.9 |
| Rank | 255 | 105 |
| Context | 131K | 200K |
| Input $/M | $0.018 | $20 |
| Output $/M | $0.09 | $80 |
| Weights | Open | Proprietary |

## Coding

- gpt-oss-20b: 37.6 (#192)
- o3-pro: 55.5 (#24)

| Benchmark | gpt-oss-20b | o3-pro |
|---|---|---|
| WeirdML | 40.9% | 58.2% |
| Aider Polyglot | — | 84.9% |
| SciCode | 34.4% | — |
| LMArena Coding | 1306 | — |
| ALE-Bench | 566.05 | — |

## Agentic & Tool Use

- gpt-oss-20b: 9.3 (#154)
- o3-pro: —

| Benchmark | gpt-oss-20b | o3-pro |
|---|---|---|
| Terminal-Bench | 3.4% | — |

## Reasoning

- gpt-oss-20b: 19.3 (#261)
- o3-pro: 23.8 (#171)

| Benchmark | gpt-oss-20b | o3-pro |
|---|---|---|
| Kagi LLM Benchmark | 53.2% | 72.1% |
| DTBench | 68% | 86.9% |
| LMCA | 14.5% | 38.5% |
| Epoch Capabilities Index | 137.82 | 147.42 |
| ARC-AGI-2 | — | 4.9% |
| ARC-AGI-1 | — | 59.3% |
| CritPt | 1.4% | — |
| Chess Puzzles | 4% | — |
| LMArena Hard Prompts | 1274 | — |

## Math

- gpt-oss-20b: 39.4 (#103)
- o3-pro: —

| Benchmark | gpt-oss-20b | o3-pro |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 65.3% | — |
| Omni-MATH | 56.5% | — |
| LMArena Math | 1317 | — |

## Knowledge

- gpt-oss-20b: 34.6 (#195)
- o3-pro: 29.5 (#238)

| Benchmark | gpt-oss-20b | o3-pro |
|---|---|---|
| GPQA Diamond | 60.8% | — |
| MMLU-Pro | 74% | — |
| Confabulations | — | 14.2% |
| Vectara Hallucination Rate | — | 23.3% |
| GPQA (HELM) | 59.4% | — |
| LMArena Expert | 1258 | — |

## Multilingual

- gpt-oss-20b: 42.2 (#197)
- o3-pro: —

| Benchmark | gpt-oss-20b | o3-pro |
|---|---|---|
| LMArena Non-English | 1268 | — |
| LMArena Chinese | 1314 | — |
| LMArena German | 1255 | — |
| LMArena Japanese | 1244 | — |
| LMArena Korean | 1236 | — |
| LMArena Russian | 1278 | — |
| LMArena Spanish | 1267 | — |

## Instruction Following

- gpt-oss-20b: 61.8 (#240)
- o3-pro: —

| Benchmark | gpt-oss-20b | o3-pro |
|---|---|---|
| IFEval | 73.2% | — |
| LMArena Instruction Following | 1236 | — |

## Long Context

- gpt-oss-20b: 37.9 (#209)
- o3-pro: 72.2 (#1)

| Benchmark | gpt-oss-20b | o3-pro |
|---|---|---|
| Fiction.LiveBench | — | 97.2% |
| LMArena Longer Query | 1250 | — |

## Writing & Preference

- gpt-oss-20b: 35.5 (#265)
- o3-pro: 57.1 (#133)

| Benchmark | gpt-oss-20b | o3-pro |
|---|---|---|
| LMArena Text | 1287 | — |
| LMArena Creative Writing | 1201 | — |
| Short-Story Creative Writing | — | 84.4% |
| EQ-Bench Creative Writing | 666 | — |
| WildBench | 73.7% | — |
| LMArena Multi-Turn | 1268 | — |

## FAQ

### Is gpt-oss-20b better than o3-pro?

o3-pro is the stronger model overall, scoring 42.9 to 32.5 on the Noometry Index. gpt-oss-20b costs 972× less per token, which makes it the better buy when o3-pro's lead doesn't matter for your workload.

### Which is cheaper, gpt-oss-20b or o3-pro?

gpt-oss-20b is cheaper. It lists at $0.018 per million input tokens and $0.09 per million output tokens; o3-pro lists at $20 and $80.

### Is gpt-oss-20b or o3-pro better for coding?

o3-pro scores higher on coding benchmarks: 55.5 versus 37.6 in the Noometry coding category.

### Which has the bigger context window?

o3-pro does, with 200K tokens against 131K.

### How many benchmarks do gpt-oss-20b and o3-pro share?

5 benchmarks have published results for both models. gpt-oss-20b has 34 scored results on Noometry and o3-pro has 12.