# GPT-5.3 Codex vs gpt-oss-20b

> GPT-5.3 Codex is the stronger model overall, scoring 45.5 to 32.1 on the Noometry Index. gpt-oss-20b costs 134× less per token, which makes it the better buy when GPT-5.3 Codex's lead doesn't matter for your workload.

- Canonical page: https://noometry.com/compare/gpt-5-3-codex-vs-gpt-oss-20b
- Last updated: 2026-10-10
- Shared benchmarks: 4

## Summary

- They share 4 benchmarks with published results for both. GPT-5.3 Codex scores higher in 2 categories and gpt-oss-20b in 0 categories; 2 gaps are clear of the uncertainty.
- The widest gap is in agentic & tool use, where GPT-5.3 Codex leads 47.8 to 9.2.
- The biggest single-benchmark swing is Terminal-Bench: 78.4% for GPT-5.3 Codex and 3.4% for gpt-oss-20b.
- gpt-oss-20b is cheaper at $0.018 / $0.09 per million input/output tokens, against $1.75 / $14 for GPT-5.3 Codex.
- GPT-5.3 Codex accepts more context: 400K tokens versus 131K.
- gpt-oss-20b has downloadable open weights; the other is API-only.

## Snapshot

| | GPT-5.3 Codex | gpt-oss-20b |
|---|---|---|
| Provider | OpenAI | OpenAI |
| Noometry Index | 45.5 | 32.1 |
| Rank | 61 | 249 |
| Context | 400K | 131K |
| Input $/M | $1.75 | $0.018 |
| Output $/M | $14 | $0.09 |
| Weights | Proprietary | Open |

## Coding

- GPT-5.3 Codex: 48.5 (#51)
- gpt-oss-20b: 37.1 (#185)

| Benchmark | GPT-5.3 Codex | gpt-oss-20b |
|---|---|---|
| WeirdML | 79.3% | 40.9% |
| ALE-Bench | 1,655 | 566.05 |
| SWE-bench Verified | 74.8% | — |
| LMArena WebDev | 1408 | — |
| SciCode | — | 34.4% |
| LMArena Coding | — | 1307 |

## Agentic & Tool Use

- GPT-5.3 Codex: 47.8 (#9)
- gpt-oss-20b: 9.2 (#146)

| Benchmark | GPT-5.3 Codex | gpt-oss-20b |
|---|---|---|
| Terminal-Bench | 78.4% | 3.4% |
| METR Time Horizons | 74.5% | — |
| Vending-Bench 2 | 5,940 | — |

## Reasoning

- GPT-5.3 Codex: —
- gpt-oss-20b: 18.8 (#251)

| Benchmark | GPT-5.3 Codex | gpt-oss-20b |
|---|---|---|
| Epoch Capabilities Index | 156.84 | 137.81 |
| Kagi LLM Benchmark | — | 53.2% |
| CritPt | — | 1.4% |
| Chess Puzzles | — | 4% |
| LMArena Hard Prompts | — | 1274 |
| DTBench | — | 68% |
| LMCA | — | 14.5% |

## Math

- GPT-5.3 Codex: —
- gpt-oss-20b: 37.6 (#93)

| Benchmark | GPT-5.3 Codex | gpt-oss-20b |
|---|---|---|
| OTIS Mock AIME 2024-2025 | — | 65.3% |
| Omni-MATH | — | 56.5% |
| LMArena Math | — | 1317 |

## Knowledge

- GPT-5.3 Codex: —
- gpt-oss-20b: 34.6 (#186)

| Benchmark | GPT-5.3 Codex | gpt-oss-20b |
|---|---|---|
| GPQA Diamond | — | 60.8% |
| MMLU-Pro | — | 74% |
| GPQA (HELM) | — | 59.4% |
| LMArena Expert | — | 1258 |

## Multilingual

- GPT-5.3 Codex: —
- gpt-oss-20b: 42.4 (#192)

| Benchmark | GPT-5.3 Codex | gpt-oss-20b |
|---|---|---|
| LMArena Non-English | — | 1269 |
| LMArena Chinese | — | 1311 |
| LMArena German | — | 1249 |
| LMArena Japanese | — | 1245 |
| LMArena Korean | — | 1239 |
| LMArena Russian | — | 1279 |
| LMArena Spanish | — | 1267 |

## Instruction Following

- GPT-5.3 Codex: —
- gpt-oss-20b: 62.0 (#235)

| Benchmark | GPT-5.3 Codex | gpt-oss-20b |
|---|---|---|
| IFEval | — | 73.2% |
| LMArena Instruction Following | — | 1236 |

## Long Context

- GPT-5.3 Codex: —
- gpt-oss-20b: 37.9 (#203)

| Benchmark | GPT-5.3 Codex | gpt-oss-20b |
|---|---|---|
| LMArena Longer Query | — | 1251 |

## Writing & Preference

- GPT-5.3 Codex: —
- gpt-oss-20b: 35.8 (#260)

| Benchmark | GPT-5.3 Codex | gpt-oss-20b |
|---|---|---|
| LMArena Text | — | 1288 |
| LMArena Creative Writing | — | 1201 |
| EQ-Bench Creative Writing | — | 666 |
| WildBench | — | 73.7% |
| LMArena Multi-Turn | — | 1268 |

## FAQ

### Is GPT-5.3 Codex better than gpt-oss-20b?

GPT-5.3 Codex is the stronger model overall, scoring 45.5 to 32.1 on the Noometry Index. gpt-oss-20b costs 134× less per token, which makes it the better buy when GPT-5.3 Codex's lead doesn't matter for your workload.

### Which is cheaper, GPT-5.3 Codex 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; GPT-5.3 Codex lists at $1.75 and $14.

### Is GPT-5.3 Codex or gpt-oss-20b better for coding?

GPT-5.3 Codex scores higher on coding benchmarks: 48.5 versus 37.1 in the Noometry coding category.

### Which has the bigger context window?

GPT-5.3 Codex does, with 400K tokens against 131K.

### How many benchmarks do GPT-5.3 Codex and gpt-oss-20b share?

4 benchmarks have published results for both models. GPT-5.3 Codex has 8 scored results on Noometry and gpt-oss-20b has 34.