Model comparison
Amazon Nova Micro vs Qwen2.5-Coder-32B
Qwen2.5-Coder-32B is the stronger model overall, scoring 33.4 to 30.4 on the Noometry Index. Amazon Nova Micro costs 12× less per token, which makes it the better buy when Qwen2.5-Coder-32B's lead doesn't matter for your workload.
Last verified . 20 shared benchmarks.
Alibaba (Qwen)Summary
- They share 20 benchmarks with published results for both. Amazon Nova Micro scores higher in 1 category and Qwen2.5-Coder-32B in 7 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in coding, where Amazon Nova Micro leads 30.5 to 22.6.
- The biggest single-benchmark swing is LiveBench Coding: 20.2% for Amazon Nova Micro and 56.9% for Qwen2.5-Coder-32B.
- Amazon Nova Micro is cheaper at $0.035 / $0.14 per million input/output tokens, against $0.66 / $1 for Qwen2.5-Coder-32B.
- Amazon Nova Micro accepts more context: 128K tokens versus 33K.
- Qwen2.5-Coder-32B has downloadable open weights; the other is API-only.
Side by side
| Amazon Nova Micro | Qwen2.5-Coder-32B | |
|---|---|---|
| Provider | Amazon | Alibaba (Qwen) |
| Noometry Index | 30.4 | 33.4 |
| Released | 2024-12-03 | 2024-09-18 |
| Weights | Proprietary | Open |
| Context window | 128K | 33K |
| Max output | 10K | 29K |
| Input $ / M tokens | $0.035 | $0.66 |
| Output $ / M tokens | $0.14 | $1 |
| Results tracked | 32 | 31 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Amazon Nova Micro leads
Amazon Nova Micro: 30.5 (#295), Qwen2.5-Coder-32B: 22.6 (#333)
| Benchmark | Amazon Nova Micro | Qwen2.5-Coder-32B |
|---|---|---|
| LiveBench Coding | 20.2% | 56.9% |
| LMArena Coding | 1218 | 1276 |
| SWE-bench Verified (bash only) | — | 9% |
| Aider Polyglot | — | 16.4% |
| BigCodeBench Instruct | — | 49% |
| BigCodeBench Complete | — | 58% |
| HumanEval+ | — | 87.2% |
| MBPP+ | — | 77% |
Agentic & Tool Use Not comparable
Amazon Nova Micro: 22.1 (#132), Qwen2.5-Coder-32B: —
| Benchmark | Amazon Nova Micro | Qwen2.5-Coder-32B |
|---|---|---|
| Berkeley Function Calling Leaderboard | 22.3% | — |
Reasoning Qwen2.5-Coder-32B leads
Amazon Nova Micro: 17.4 (#294), Qwen2.5-Coder-32B: 21.2 (#225)
| Benchmark | Amazon Nova Micro | Qwen2.5-Coder-32B |
|---|---|---|
| LiveBench Reasoning | 25.1% | 42.1% |
| LMArena Hard Prompts | 1191 | 1251 |
| LiveBench Data Analysis | 34% | 49.9% |
| LiveBench | 29.6% | 46.2% |
| Epoch Capabilities Index | — | 119.49 |
| HellaSwag | — | 83% |
| WinoGrande | — | 80.8% |
Math Qwen2.5-Coder-32B leads
Amazon Nova Micro: 26.9 (#254), Qwen2.5-Coder-32B: 33.3 (#204)
| Benchmark | Amazon Nova Micro | Qwen2.5-Coder-32B |
|---|---|---|
| LiveBench Math | 34.5% | 46.6% |
| LMArena Math | 1206 | 1251 |
| Omni-MATH | 21.4% | — |
| GSM8K | — | 93% |
Knowledge Qwen2.5-Coder-32B leads
Amazon Nova Micro: 29.6 (#237), Qwen2.5-Coder-32B: 33.4 (#203)
| Benchmark | Amazon Nova Micro | Qwen2.5-Coder-32B |
|---|---|---|
| LMArena Expert | 1184 | 1221 |
| MMLU | 70.8% | 79.1% |
| MMLU-Pro | 51.1% | — |
| Vectara Hallucination Rate | 5.5% | — |
| GPQA (HELM) | 38.3% | — |
| ARC (AI2) Challenge | — | 70.5% |
Multilingual Qwen2.5-Coder-32B leads
Amazon Nova Micro: 36.5 (#239), Qwen2.5-Coder-32B: 37.8 (#235)
| Benchmark | Amazon Nova Micro | Qwen2.5-Coder-32B |
|---|---|---|
| LMArena Non-English | 1186 | 1205 |
| LMArena Chinese | 1209 | 1222 |
| LMArena Russian | 1185 | 1228 |
| LMArena French | 1238 | — |
| LMArena German | 1192 | — |
| LMArena Japanese | 1154 | — |
| LMArena Korean | 1150 | — |
| LMArena Spanish | 1225 | — |
Instruction Following Qwen2.5-Coder-32B leads
Amazon Nova Micro: 56.3 (#272), Qwen2.5-Coder-32B: 61.4 (#245)
| Benchmark | Amazon Nova Micro | Qwen2.5-Coder-32B |
|---|---|---|
| LiveBench Instruction Following | 48% | 58.7% |
| LMArena Instruction Following | 1174 | 1223 |
| IFEval | 76% | — |
Long Context Qwen2.5-Coder-32B leads
Amazon Nova Micro: 36.5 (#229), Qwen2.5-Coder-32B: 38.0 (#208)
| Benchmark | Amazon Nova Micro | Qwen2.5-Coder-32B |
|---|---|---|
| LMArena Longer Query | 1205 | 1251 |
Writing & Preference Qwen2.5-Coder-32B leads
Amazon Nova Micro: 39.5 (#247), Qwen2.5-Coder-32B: 41.6 (#240)
| Benchmark | Amazon Nova Micro | Qwen2.5-Coder-32B |
|---|---|---|
| LMArena Text | 1208 | 1230 |
| LMArena Creative Writing | 1172 | 1174 |
| LMArena Multi-Turn | 1178 | 1222 |
| LiveBench Language | 15.8% | 23.3% |
| WildBench | 74.3% | — |
Frequently asked questions
Is Amazon Nova Micro better than Qwen2.5-Coder-32B?
Qwen2.5-Coder-32B is the stronger model overall, scoring 33.4 to 30.4 on the Noometry Index. Amazon Nova Micro costs 12× less per token, which makes it the better buy when Qwen2.5-Coder-32B's lead doesn't matter for your workload.
Which is cheaper, Amazon Nova Micro or Qwen2.5-Coder-32B?
Amazon Nova Micro is cheaper. It lists at $0.035 per million input tokens and $0.14 per million output tokens; Qwen2.5-Coder-32B lists at $0.66 and $1.
Is Amazon Nova Micro or Qwen2.5-Coder-32B better for coding?
Amazon Nova Micro scores higher on coding benchmarks: 30.5 versus 22.6 in the Noometry coding category.
Which has the bigger context window?
Amazon Nova Micro does, with 128K tokens against 33K.
How many benchmarks do Amazon Nova Micro and Qwen2.5-Coder-32B share?
20 benchmarks have published results for both models. Amazon Nova Micro has 32 scored results on Noometry and Qwen2.5-Coder-32B has 31.