# Devstral Small 2505 vs Llama 3.2 3B

> Devstral Small 2505 is the stronger model overall, scoring 33.9 to 28.4 on the Noometry Index.

- Canonical page: https://noometry.com/compare/devstral-small-vs-llama-3-2-3b
- Last updated: 2026-10-10
- Shared benchmarks: 0

## Summary

- The widest gap is in coding, where Devstral Small 2505 leads 38.6 to 27.1.
- Llama 3.2 3B is cheaper at $0.05 / $0.33 per million input/output tokens, against $0.10 / $0.30 for Devstral Small 2505.
- Llama 3.2 3B accepts more context: 131K tokens versus 128K.

## Snapshot

| | Devstral Small 2505 | Llama 3.2 3B |
|---|---|---|
| Provider | Mistral AI | Meta |
| Noometry Index | 33.9 | 28.4 |
| Rank | 228 | 318 |
| Context | 128K | 131K |
| Input $/M | $0.10 | $0.05 |
| Output $/M | $0.30 | $0.33 |
| Weights | Open | Open |

## Coding

- Devstral Small 2505: 38.6 (#160)
- Llama 3.2 3B: 27.1 (#314)

| Benchmark | Devstral Small 2505 | Llama 3.2 3B |
|---|---|---|
| SWE-bench Verified (bash only) | 56.4% | — |
| SciCode | 28.8% | — |
| BigCodeBench Instruct | — | 23.4% |
| LMArena Coding | — | 1098 |
| BigCodeBench Complete | — | 28.3% |

## Agentic & Tool Use

- Devstral Small 2505: —
- Llama 3.2 3B: 19.9 (#135)

| Benchmark | Devstral Small 2505 | Llama 3.2 3B |
|---|---|---|
| Berkeley Function Calling Leaderboard | — | 21.9% |
| BALROG | — | 10.1% |

## Reasoning

- Devstral Small 2505: 19.1 (#243)
- Llama 3.2 3B: 20.1 (#225)

| Benchmark | Devstral Small 2505 | Llama 3.2 3B |
|---|---|---|
| Kagi LLM Benchmark | 37.7% | — |
| CritPt | 0% | — |
| LMArena Hard Prompts | — | 1095 |

## Math

- Devstral Small 2505: —
- Llama 3.2 3B: 29.9 (#213)

| Benchmark | Devstral Small 2505 | Llama 3.2 3B |
|---|---|---|
| LMArena Math | — | 1126 |

## Knowledge

- Devstral Small 2505: —
- Llama 3.2 3B: 29.1 (#232)

| Benchmark | Devstral Small 2505 | Llama 3.2 3B |
|---|---|---|
| LMArena Expert | — | 1090 |

## Multilingual

- Devstral Small 2505: —
- Llama 3.2 3B: 26.4 (#276)

| Benchmark | Devstral Small 2505 | Llama 3.2 3B |
|---|---|---|
| LMArena Non-English | — | 1019 |
| LMArena Chinese | — | 1017 |
| LMArena German | — | 1056 |
| LMArena Russian | — | 949 |

## Instruction Following

- Devstral Small 2505: —
- Llama 3.2 3B: 56.3 (#270)

| Benchmark | Devstral Small 2505 | Llama 3.2 3B |
|---|---|---|
| LMArena Instruction Following | — | 1089 |

## Long Context

- Devstral Small 2505: —
- Llama 3.2 3B: 33.4 (#256)

| Benchmark | Devstral Small 2505 | Llama 3.2 3B |
|---|---|---|
| LMArena Longer Query | — | 1101 |

## Writing & Preference

- Devstral Small 2505: —
- Llama 3.2 3B: 24.9 (#302)

| Benchmark | Devstral Small 2505 | Llama 3.2 3B |
|---|---|---|
| LMArena Text | — | 1110 |
| LMArena Creative Writing | — | 1094 |
| EQ-Bench Creative Writing | — | 595 |
| LMArena Multi-Turn | — | 1105 |

## FAQ

### Is Devstral Small 2505 better than Llama 3.2 3B?

Devstral Small 2505 is the stronger model overall, scoring 33.9 to 28.4 on the Noometry Index.

### Which is cheaper, Devstral Small 2505 or Llama 3.2 3B?

Llama 3.2 3B is cheaper. It lists at $0.05 per million input tokens and $0.33 per million output tokens; Devstral Small 2505 lists at $0.10 and $0.30.

### Is Devstral Small 2505 or Llama 3.2 3B better for coding?

Devstral Small 2505 scores higher on coding benchmarks: 38.6 versus 27.1 in the Noometry coding category.

### Which has the bigger context window?

Llama 3.2 3B does, with 131K tokens against 128K.

### How many benchmarks do Devstral Small 2505 and Llama 3.2 3B share?

0 benchmarks have published results for both models. Devstral Small 2505 has 4 scored results on Noometry and Llama 3.2 3B has 18.