# Granite 4.0 Micro vs Qwen3 32B

> Qwen3 32B is the stronger model overall, scoring 38.8 to 28.8 on the Noometry Index. Granite 4.0 Micro costs 30× 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/granite-4-0-micro-vs-qwen3-32b
- Last updated: 2026-10-10
- Shared benchmarks: 3

## Summary

- They share 3 benchmarks with published results for both. Granite 4.0 Micro scores higher in 1 category and Qwen3 32B in 4 categories; 4 gaps are clear of the uncertainty.
- The widest gap is in knowledge, where Qwen3 32B leads 40.0 to 9.9.
- The biggest single-benchmark swing is OTIS Mock AIME 2024-2025: 2.8% for Granite 4.0 Micro and 66.9% for Qwen3 32B.
- Granite 4.0 Micro is cheaper at $0.017 / $0.11 per million input/output tokens, against $0.70 / $2.80 for Qwen3 32B.
- Qwen3 32B accepts more context: 131K tokens versus 131K.

## Snapshot

| | Granite 4.0 Micro | Qwen3 32B |
|---|---|---|
| Provider | IBM | Alibaba (Qwen) |
| Noometry Index | 28.8 | 38.8 |
| Rank | 312 | 166 |
| Context | 131K | 131K |
| Input $/M | $0.017 | $0.70 |
| Output $/M | $0.11 | $2.80 |
| Weights | Open | Open |

## Coding

- Granite 4.0 Micro: —
- Qwen3 32B: 37.3 (#184)

| Benchmark | Granite 4.0 Micro | Qwen3 32B |
|---|---|---|
| Aider Polyglot | — | 40% |
| SciCode | — | 35.4% |
| LMArena Coding | — | 1358 |

## Agentic & Tool Use

- Granite 4.0 Micro: —
- Qwen3 32B: 32.3 (#58)

| Benchmark | Granite 4.0 Micro | Qwen3 32B |
|---|---|---|
| Berkeley Function Calling Leaderboard | — | 48.7% |

## Reasoning

- Granite 4.0 Micro: 18.5 (#261)
- Qwen3 32B: 19.6 (#233)

| Benchmark | Granite 4.0 Micro | Qwen3 32B |
|---|---|---|
| Chess Puzzles | 0% | 5% |
| Kagi LLM Benchmark | — | 54.9% |
| CritPt | — | 0.3% |
| LMArena Hard Prompts | — | 1335 |
| DTBench | — | 67.5% |
| LMCA | — | 17.3% |
| Epoch Capabilities Index | — | 138.51 |

## Math

- Granite 4.0 Micro: 11.2 (#301)
- Qwen3 32B: 37.9 (#89)

| Benchmark | Granite 4.0 Micro | Qwen3 32B |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 2.8% | 66.9% |
| Omni-MATH | 20.9% | — |
| LMArena Math | — | 1399 |

## Knowledge

- Granite 4.0 Micro: 9.9 (#299)
- Qwen3 32B: 40.0 (#112)

| Benchmark | Granite 4.0 Micro | Qwen3 32B |
|---|---|---|
| GPQA Diamond | 28.3% | 65.7% |
| MMLU-Pro | 39.5% | — |
| Vectara Hallucination Rate | — | 5.9% |
| GPQA (HELM) | 30.7% | — |
| LMArena Expert | — | 1362 |

## Multilingual

- Granite 4.0 Micro: —
- Qwen3 32B: 45.8 (#162)

| Benchmark | Granite 4.0 Micro | Qwen3 32B |
|---|---|---|
| LMArena Non-English | — | 1317 |
| LMArena Chinese | — | 1357 |
| LMArena German | — | 1341 |
| LMArena Russian | — | 1311 |

## Instruction Following

- Granite 4.0 Micro: 70.0 (#164)
- Qwen3 32B: 69.1 (#174)

| Benchmark | Granite 4.0 Micro | Qwen3 32B |
|---|---|---|
| IFEval | 84.9% | — |
| LMArena Instruction Following | — | 1305 |

## Long Context

- Granite 4.0 Micro: —
- Qwen3 32B: 43.7 (#83)

| Benchmark | Granite 4.0 Micro | Qwen3 32B |
|---|---|---|
| Fiction.LiveBench | — | 74.2% |
| LMArena Longer Query | — | 1327 |

## Writing & Preference

- Granite 4.0 Micro: 46.7 (#212)
- Qwen3 32B: 53.1 (#158)

| Benchmark | Granite 4.0 Micro | Qwen3 32B |
|---|---|---|
| LMArena Text | — | 1340 |
| LMArena Creative Writing | — | 1297 |
| WildBench | 67% | — |
| LMArena Multi-Turn | — | 1331 |

## FAQ

### Is Granite 4.0 Micro better than Qwen3 32B?

Qwen3 32B is the stronger model overall, scoring 38.8 to 28.8 on the Noometry Index. Granite 4.0 Micro costs 30× less per token, which makes it the better buy when Qwen3 32B's lead doesn't matter for your workload.

### Which is cheaper, Granite 4.0 Micro or Qwen3 32B?

Granite 4.0 Micro is cheaper. It lists at $0.017 per million input tokens and $0.11 per million output tokens; Qwen3 32B lists at $0.70 and $2.80.

### Which has the bigger context window?

Qwen3 32B does, with 131K tokens against 131K.

### How many benchmarks do Granite 4.0 Micro and Qwen3 32B share?

3 benchmarks have published results for both models. Granite 4.0 Micro has 8 scored results on Noometry and Qwen3 32B has 26.