# Devstral Small 2505 vs gpt-oss-20b

> Devstral Small 2505 is the stronger model overall, scoring 33.9 to 32.1 on the Noometry Index. gpt-oss-20b costs 4.2× less per token, which makes it the better buy when Devstral Small 2505's lead doesn't matter for your workload.

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

## Summary

- They share 3 benchmarks with published results for both. Devstral Small 2505 scores higher in 2 categories and gpt-oss-20b in 0 categories; one gap is clear of the uncertainty.
- The biggest single-benchmark swing is Kagi LLM Benchmark: 37.7% for Devstral Small 2505 and 53.2% for gpt-oss-20b.
- gpt-oss-20b is cheaper at $0.018 / $0.09 per million input/output tokens, against $0.10 / $0.30 for Devstral Small 2505.
- gpt-oss-20b accepts more context: 131K tokens versus 128K.

## Snapshot

| | Devstral Small 2505 | gpt-oss-20b |
|---|---|---|
| Provider | Mistral AI | OpenAI |
| Noometry Index | 33.9 | 32.1 |
| Rank | 228 | 249 |
| Context | 128K | 131K |
| Input $/M | $0.10 | $0.018 |
| Output $/M | $0.30 | $0.09 |
| Weights | Open | Open |

## Coding

- Devstral Small 2505: 38.6 (#160)
- gpt-oss-20b: 37.1 (#185)

| Benchmark | Devstral Small 2505 | gpt-oss-20b |
|---|---|---|
| SciCode | 28.8% | 34.4% |
| SWE-bench Verified (bash only) | 56.4% | — |
| WeirdML | — | 40.9% |
| LMArena Coding | — | 1307 |
| ALE-Bench | — | 566.05 |

## Agentic & Tool Use

- Devstral Small 2505: —
- gpt-oss-20b: 9.2 (#146)

| Benchmark | Devstral Small 2505 | gpt-oss-20b |
|---|---|---|
| Terminal-Bench | — | 3.4% |

## Reasoning

- Devstral Small 2505: 19.1 (#243)
- gpt-oss-20b: 18.8 (#251)

| Benchmark | Devstral Small 2505 | gpt-oss-20b |
|---|---|---|
| Kagi LLM Benchmark | 37.7% | 53.2% |
| CritPt | 0% | 1.4% |
| Chess Puzzles | — | 4% |
| LMArena Hard Prompts | — | 1274 |
| DTBench | — | 68% |
| LMCA | — | 14.5% |
| Epoch Capabilities Index | — | 137.81 |

## Math

- Devstral Small 2505: —
- gpt-oss-20b: 37.6 (#93)

| Benchmark | Devstral Small 2505 | gpt-oss-20b |
|---|---|---|
| OTIS Mock AIME 2024-2025 | — | 65.3% |
| Omni-MATH | — | 56.5% |
| LMArena Math | — | 1317 |

## Knowledge

- Devstral Small 2505: —
- gpt-oss-20b: 34.6 (#186)

| Benchmark | Devstral Small 2505 | gpt-oss-20b |
|---|---|---|
| GPQA Diamond | — | 60.8% |
| MMLU-Pro | — | 74% |
| GPQA (HELM) | — | 59.4% |
| LMArena Expert | — | 1258 |

## Multilingual

- Devstral Small 2505: —
- gpt-oss-20b: 42.4 (#192)

| Benchmark | Devstral Small 2505 | 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

- Devstral Small 2505: —
- gpt-oss-20b: 62.0 (#235)

| Benchmark | Devstral Small 2505 | gpt-oss-20b |
|---|---|---|
| IFEval | — | 73.2% |
| LMArena Instruction Following | — | 1236 |

## Long Context

- Devstral Small 2505: —
- gpt-oss-20b: 37.9 (#203)

| Benchmark | Devstral Small 2505 | gpt-oss-20b |
|---|---|---|
| LMArena Longer Query | — | 1251 |

## Writing & Preference

- Devstral Small 2505: —
- gpt-oss-20b: 35.8 (#260)

| Benchmark | Devstral Small 2505 | gpt-oss-20b |
|---|---|---|
| LMArena Text | — | 1288 |
| LMArena Creative Writing | — | 1201 |
| EQ-Bench Creative Writing | — | 666 |
| WildBench | — | 73.7% |
| LMArena Multi-Turn | — | 1268 |

## FAQ

### Is Devstral Small 2505 better than gpt-oss-20b?

Devstral Small 2505 is the stronger model overall, scoring 33.9 to 32.1 on the Noometry Index. gpt-oss-20b costs 4.2× less per token, which makes it the better buy when Devstral Small 2505's lead doesn't matter for your workload.

### Which is cheaper, Devstral Small 2505 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; Devstral Small 2505 lists at $0.10 and $0.30.

### Is Devstral Small 2505 or gpt-oss-20b better for coding?

Devstral Small 2505 scores higher on coding benchmarks: 38.6 versus 37.1 in the Noometry coding category.

### Which has the bigger context window?

gpt-oss-20b does, with 131K tokens against 128K.

### How many benchmarks do Devstral Small 2505 and gpt-oss-20b share?

3 benchmarks have published results for both models. Devstral Small 2505 has 4 scored results on Noometry and gpt-oss-20b has 34.