# Amazon Nova Lite vs GPT-4

> Amazon Nova Lite is the stronger model overall, scoring 31.9 to 29.1 on the Noometry Index.

- Canonical page: https://noometry.com/compare/amazon-nova-lite-vs-gpt-4
- Last updated: 2026-10-12
- Shared benchmarks: 18

## Summary

- They share 18 benchmarks with published results for both. Amazon Nova Lite scores higher in 5 categories and GPT-4 in 3 categories; 6 gaps are clear of the uncertainty.
- The widest gap is in math, where Amazon Nova Lite leads 27.8 to 10.8.
- Amazon Nova Lite is cheaper at $0.06 / $0.24 per million input/output tokens, against $30 / $60 for GPT-4.
- Amazon Nova Lite accepts more context: 300K tokens versus 8K.

## Snapshot

| | Amazon Nova Lite | GPT-4 |
|---|---|---|
| Provider | Amazon | OpenAI |
| Noometry Index | 31.9 | 29.1 |
| Rank | 265 | 316 |
| Context | 300K | 8K |
| Input $/M | $0.06 | $30 |
| Output $/M | $0.24 | $60 |
| Weights | Proprietary | Proprietary |

## Coding

- Amazon Nova Lite: 32.5 (#270)
- GPT-4: 31.6 (#283)

| Benchmark | Amazon Nova Lite | GPT-4 |
|---|---|---|
| LMArena Coding | 1239 | 1254 |
| WeirdML | — | 12.4% |
| BigCodeBench Instruct | — | 46% |
| LiveBench Coding | 27.5% | — |
| BigCodeBench Complete | — | 57.2% |
| ALE-Bench | 236.25 | — |
| HumanEval+ | — | 79.3% |

## Agentic & Tool Use

- Amazon Nova Lite: —
- GPT-4: —

| Benchmark | Amazon Nova Lite | GPT-4 |
|---|---|---|
| METR Time Horizons | — | 36.1% |

## Reasoning

- Amazon Nova Lite: 19.3 (#260)
- GPT-4: 17.8 (#289)

| Benchmark | Amazon Nova Lite | GPT-4 |
|---|---|---|
| LMArena Hard Prompts | 1220 | 1241 |
| Chess Puzzles | — | 4% |
| LiveBench Reasoning | 36.7% | — |
| Mystery Game Puzzles | — | 12% |
| DTBench | — | 62.7% |
| LiveBench Data Analysis | 37.2% | — |
| LMCA | — | 17.1% |
| BIG-Bench Hard | — | 75.1% |
| Epoch Capabilities Index | — | 125.89 |
| ForecastBench | — | 57.8 |
| HellaSwag | — | 95.3% |
| LiveBench | 36.4% | — |
| WinoGrande | — | 87.5% |

## Math

- Amazon Nova Lite: 27.8 (#247)
- GPT-4: 10.8 (#309)

| Benchmark | Amazon Nova Lite | GPT-4 |
|---|---|---|
| LMArena Math | 1227 | 1269 |
| OTIS Mock AIME 2024-2025 | — | 1.1% |
| Omni-MATH | 23.3% | — |
| LiveBench Math | 36.7% | — |
| MATH Level 5 | — | 23% |
| GSM8K | — | 92% |

## Knowledge

- Amazon Nova Lite: 26.3 (#259)
- GPT-4: 18.4 (#282)

| Benchmark | Amazon Nova Lite | GPT-4 |
|---|---|---|
| LMArena Expert | 1201 | 1211 |
| MMLU | 77% | 86.4% |
| GPQA Diamond | — | 35.7% |
| Humanity's Last Exam | 3.6% | — |
| MMLU-Pro | 60% | — |
| Vectara Hallucination Rate | 6.1% | — |
| GPQA (HELM) | 39.7% | — |
| TriviaQA | — | 84.8% |

## Multimodal

- Amazon Nova Lite: 25.5 (#123)
- GPT-4: —

| Benchmark | Amazon Nova Lite | GPT-4 |
|---|---|---|
| LMArena Vision | 990 | — |

## Multilingual

- Amazon Nova Lite: 38.0 (#232)
- GPT-4: 40.6 (#215)

| Benchmark | Amazon Nova Lite | GPT-4 |
|---|---|---|
| LMArena Non-English | 1208 | 1246 |
| LMArena Chinese | 1225 | 1242 |
| LMArena French | 1237 | 1283 |
| LMArena German | 1229 | 1251 |
| LMArena Japanese | 1153 | 1209 |
| LMArena Korean | 1153 | 1184 |
| LMArena Russian | 1216 | 1251 |
| LMArena Spanish | 1226 | 1261 |

## Instruction Following

- Amazon Nova Lite: 59.3 (#255)
- GPT-4: 65.3 (#222)

| Benchmark | Amazon Nova Lite | GPT-4 |
|---|---|---|
| LMArena Instruction Following | 1205 | 1241 |
| LiveBench Instruction Following | 54.1% | — |
| IFEval | 77.6% | — |

## Long Context

- Amazon Nova Lite: 37.4 (#217)
- GPT-4: 37.7 (#212)

| Benchmark | Amazon Nova Lite | GPT-4 |
|---|---|---|
| LMArena Longer Query | 1234 | 1244 |

## Writing & Preference

- Amazon Nova Lite: 42.1 (#237)
- GPT-4: 34.9 (#268)

| Benchmark | Amazon Nova Lite | GPT-4 |
|---|---|---|
| LMArena Text | 1229 | 1263 |
| LMArena Creative Writing | 1197 | 1244 |
| LMArena Multi-Turn | 1199 | 1257 |
| EQ-Bench Creative Writing | — | 752 |
| WildBench | 75% | — |
| LiveBench Language | 25.9% | — |

## FAQ

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

Amazon Nova Lite is the stronger model overall, scoring 31.9 to 29.1 on the Noometry Index.

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

Amazon Nova Lite is cheaper. It lists at $0.06 per million input tokens and $0.24 per million output tokens; GPT-4 lists at $30 and $60.

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

They score almost the same on coding (32.5 vs 31.6); test both on your own repository before choosing.

### Which has the bigger context window?

Amazon Nova Lite does, with 300K tokens against 8K.

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

18 benchmarks have published results for both models. Amazon Nova Lite has 34 scored results on Noometry and GPT-4 has 38.