# Amazon Nova Micro vs GPT-4

> Amazon Nova Micro is the stronger model overall, scoring 30.4 to 29.1 on the Noometry Index.

- Canonical page: https://noometry.com/compare/amazon-nova-micro-vs-gpt-4
- Last updated: 2026-10-11
- Shared benchmarks: 18

## Summary

- They share 18 benchmarks with published results for both. Amazon Nova Micro scores higher in 3 categories and GPT-4 in 5 categories; 7 gaps are clear of the uncertainty.
- The widest gap is in math, where Amazon Nova Micro leads 26.9 to 10.8.
- Amazon Nova Micro is cheaper at $0.035 / $0.14 per million input/output tokens, against $30 / $60 for GPT-4.
- Amazon Nova Micro accepts more context: 128K tokens versus 8K.

## Snapshot

| | Amazon Nova Micro | GPT-4 |
|---|---|---|
| Provider | Amazon | OpenAI |
| Noometry Index | 30.4 | 29.1 |
| Rank | 294 | 316 |
| Context | 128K | 8K |
| Input $/M | $0.035 | $30 |
| Output $/M | $0.14 | $60 |
| Weights | Proprietary | Proprietary |

## Coding

- Amazon Nova Micro: 30.5 (#295)
- GPT-4: 31.6 (#283)

| Benchmark | Amazon Nova Micro | GPT-4 |
|---|---|---|
| LMArena Coding | 1218 | 1254 |
| WeirdML | — | 12.4% |
| BigCodeBench Instruct | — | 46% |
| LiveBench Coding | 20.2% | — |
| BigCodeBench Complete | — | 57.2% |
| HumanEval+ | — | 79.3% |

## Agentic & Tool Use

- Amazon Nova Micro: 22.1 (#132)
- GPT-4: —

| Benchmark | Amazon Nova Micro | GPT-4 |
|---|---|---|
| Berkeley Function Calling Leaderboard | 22.3% | — |
| METR Time Horizons | — | 36.1% |

## Reasoning

- Amazon Nova Micro: 17.4 (#294)
- GPT-4: 17.8 (#289)

| Benchmark | Amazon Nova Micro | GPT-4 |
|---|---|---|
| LMArena Hard Prompts | 1191 | 1241 |
| Chess Puzzles | — | 4% |
| LiveBench Reasoning | 25.1% | — |
| Mystery Game Puzzles | — | 12% |
| DTBench | — | 62.7% |
| LiveBench Data Analysis | 34% | — |
| LMCA | — | 17.1% |
| BIG-Bench Hard | — | 75.1% |
| Epoch Capabilities Index | — | 125.89 |
| ForecastBench | — | 57.8 |
| HellaSwag | — | 95.3% |
| LiveBench | 29.6% | — |
| WinoGrande | — | 87.5% |

## Math

- Amazon Nova Micro: 26.9 (#254)
- GPT-4: 10.8 (#309)

| Benchmark | Amazon Nova Micro | GPT-4 |
|---|---|---|
| LMArena Math | 1206 | 1269 |
| OTIS Mock AIME 2024-2025 | — | 1.1% |
| Omni-MATH | 21.4% | — |
| LiveBench Math | 34.5% | — |
| MATH Level 5 | — | 23% |
| GSM8K | — | 92% |

## Knowledge

- Amazon Nova Micro: 29.6 (#237)
- GPT-4: 18.4 (#282)

| Benchmark | Amazon Nova Micro | GPT-4 |
|---|---|---|
| LMArena Expert | 1184 | 1211 |
| MMLU | 70.8% | 86.4% |
| GPQA Diamond | — | 35.7% |
| MMLU-Pro | 51.1% | — |
| Vectara Hallucination Rate | 5.5% | — |
| GPQA (HELM) | 38.3% | — |
| TriviaQA | — | 84.8% |

## Multilingual

- Amazon Nova Micro: 36.5 (#239)
- GPT-4: 40.6 (#215)

| Benchmark | Amazon Nova Micro | GPT-4 |
|---|---|---|
| LMArena Non-English | 1186 | 1246 |
| LMArena Chinese | 1209 | 1242 |
| LMArena French | 1238 | 1283 |
| LMArena German | 1192 | 1251 |
| LMArena Japanese | 1154 | 1209 |
| LMArena Korean | 1150 | 1184 |
| LMArena Russian | 1185 | 1251 |
| LMArena Spanish | 1225 | 1261 |

## Instruction Following

- Amazon Nova Micro: 56.3 (#272)
- GPT-4: 65.3 (#222)

| Benchmark | Amazon Nova Micro | GPT-4 |
|---|---|---|
| LMArena Instruction Following | 1174 | 1241 |
| LiveBench Instruction Following | 48% | — |
| IFEval | 76% | — |

## Long Context

- Amazon Nova Micro: 36.5 (#229)
- GPT-4: 37.7 (#212)

| Benchmark | Amazon Nova Micro | GPT-4 |
|---|---|---|
| LMArena Longer Query | 1205 | 1244 |

## Writing & Preference

- Amazon Nova Micro: 39.5 (#247)
- GPT-4: 34.9 (#268)

| Benchmark | Amazon Nova Micro | GPT-4 |
|---|---|---|
| LMArena Text | 1208 | 1263 |
| LMArena Creative Writing | 1172 | 1244 |
| LMArena Multi-Turn | 1178 | 1257 |
| EQ-Bench Creative Writing | — | 752 |
| WildBench | 74.3% | — |
| LiveBench Language | 15.8% | — |

## FAQ

### Is Amazon Nova Micro better than GPT-4?

Amazon Nova Micro is the stronger model overall, scoring 30.4 to 29.1 on the Noometry Index.

### Which is cheaper, Amazon Nova Micro or GPT-4?

Amazon Nova Micro is cheaper. It lists at $0.035 per million input tokens and $0.14 per million output tokens; GPT-4 lists at $30 and $60.

### Is Amazon Nova Micro or GPT-4 better for coding?

GPT-4 scores higher on coding benchmarks: 31.6 versus 30.5 in the Noometry coding category.

### Which has the bigger context window?

Amazon Nova Micro does, with 128K tokens against 8K.

### How many benchmarks do Amazon Nova Micro and GPT-4 share?

18 benchmarks have published results for both models. Amazon Nova Micro has 32 scored results on Noometry and GPT-4 has 38.