# Qwen3 32B vs Qwen3 8B

> Qwen3 32B is the stronger model overall, scoring 39.2 to 33.7 on the Noometry Index. Qwen3 8B costs 4.0× less per token, which makes it the better buy when Qwen3 32B's lead doesn't matter for your workload.

- Canonical page: https://noometry.com/compare/qwen3-32b-vs-qwen3-8b
- Last updated: 2026-10-12
- Shared benchmarks: 11

## Summary

- They share 11 benchmarks with published results for both. Qwen3 32B scores higher in 6 categories and Qwen3 8B in 0 categories; 6 gaps are clear of the uncertainty.
- The widest gap is in long context, where Qwen3 32B leads 43.8 to 37.9.
- The biggest single-benchmark swing is SciCode: 35.4% for Qwen3 32B and 22.6% for Qwen3 8B.
- Qwen3 8B is cheaper at $0.18 / $0.70 per million input/output tokens, against $0.70 / $2.80 for Qwen3 32B.

## Snapshot

| | Qwen3 32B | Qwen3 8B |
|---|---|---|
| Provider | Alibaba (Qwen) | Alibaba (Qwen) |
| Noometry Index | 39.2 | 33.7 |
| Rank | 172 | 238 |
| Context | 131K | 131K |
| Input $/M | $0.70 | $0.18 |
| Output $/M | $2.80 | $0.70 |
| Weights | Open | Open |

## Coding

- Qwen3 32B: 37.7 (#190)
- Qwen3 8B: 34.0 (#248)

| Benchmark | Qwen3 32B | Qwen3 8B |
|---|---|---|
| SciCode | 35.4% | 22.6% |
| Aider Polyglot | 40% | — |
| LMArena Coding | 1358 | — |

## Agentic & Tool Use

- Qwen3 32B: 32.6 (#62)
- Qwen3 8B: 30.2 (#78)

| Benchmark | Qwen3 32B | Qwen3 8B |
|---|---|---|
| Berkeley Function Calling Leaderboard | 48.7% | 42.6% |

## Reasoning

- Qwen3 32B: 20.2 (#241)
- Qwen3 8B: 16.6 (#303)

| Benchmark | Qwen3 32B | Qwen3 8B |
|---|---|---|
| CritPt | 0.3% | 0% |
| Chess Puzzles | 5% | 5% |
| DTBench | 67.5% | 59.7% |
| LMCA | 17.3% | 8.8% |
| Epoch Capabilities Index | 138.51 | 136.17 |
| Kagi LLM Benchmark | 54.9% | — |
| LMArena Hard Prompts | 1334 | — |

## Math

- Qwen3 32B: 39.7 (#99)
- Qwen3 8B: 34.9 (#191)

| Benchmark | Qwen3 32B | Qwen3 8B |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 66.9% | 56.1% |
| LMArena Math | 1399 | — |

## Knowledge

- Qwen3 32B: 40.0 (#125)
- Qwen3 8B: 36.1 (#173)

| Benchmark | Qwen3 32B | Qwen3 8B |
|---|---|---|
| GPQA Diamond | 65.7% | 56.8% |
| Vectara Hallucination Rate | 5.9% | 4.8% |
| LMArena Expert | 1362 | — |

## Multilingual

- Qwen3 32B: 45.6 (#167)
- Qwen3 8B: —

| Benchmark | Qwen3 32B | Qwen3 8B |
|---|---|---|
| LMArena Non-English | 1317 | — |
| LMArena Chinese | 1357 | — |
| LMArena German | 1341 | — |
| LMArena Russian | 1311 | — |

## Instruction Following

- Qwen3 32B: 68.9 (#179)
- Qwen3 8B: —

| Benchmark | Qwen3 32B | Qwen3 8B |
|---|---|---|
| LMArena Instruction Following | 1305 | — |

## Long Context

- Qwen3 32B: 43.8 (#87)
- Qwen3 8B: 37.9 (#210)

| Benchmark | Qwen3 32B | Qwen3 8B |
|---|---|---|
| Fiction.LiveBench | 74.2% | 62.1% |
| LMArena Longer Query | 1327 | — |

## Writing & Preference

- Qwen3 32B: 52.9 (#163)
- Qwen3 8B: —

| Benchmark | Qwen3 32B | Qwen3 8B |
|---|---|---|
| LMArena Text | 1340 | — |
| LMArena Creative Writing | 1297 | — |
| LMArena Multi-Turn | 1331 | — |

## FAQ

### Is Qwen3 32B better than Qwen3 8B?

Qwen3 32B is the stronger model overall, scoring 39.2 to 33.7 on the Noometry Index. Qwen3 8B costs 4.0× less per token, which makes it the better buy when Qwen3 32B's lead doesn't matter for your workload.

### Which is cheaper, Qwen3 32B or Qwen3 8B?

Qwen3 8B is cheaper. It lists at $0.18 per million input tokens and $0.70 per million output tokens; Qwen3 32B lists at $0.70 and $2.80.

### Is Qwen3 32B or Qwen3 8B better for coding?

Qwen3 32B scores higher on coding benchmarks: 37.7 versus 34.0 in the Noometry coding category.

### Which has the bigger context window?

Both accept 131K tokens.

### How many benchmarks do Qwen3 32B and Qwen3 8B share?

11 benchmarks have published results for both models. Qwen3 32B has 26 scored results on Noometry and Qwen3 8B has 11.