# Amazon Nova Lite vs GPT-4o mini

> Amazon Nova Lite is the stronger model overall, scoring 31.9 to 25.5 on the Noometry Index.

- Canonical page: https://noometry.com/compare/amazon-nova-lite-vs-gpt-4o-mini
- Last updated: 2026-10-11
- Shared benchmarks: 31

## Summary

- They share 31 benchmarks with published results for both. Amazon Nova Lite scores higher in 5 categories and GPT-4o mini in 4 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in math, where Amazon Nova Lite leads 27.8 to 10.4.
- The biggest single-benchmark swing is LiveBench Coding: 27.5% for Amazon Nova Lite and 43.1% for GPT-4o mini.
- Amazon Nova Lite is cheaper at $0.06 / $0.24 per million input/output tokens, against $0.15 / $0.60 for GPT-4o mini.
- Amazon Nova Lite accepts more context: 300K tokens versus 128K.

## Snapshot

| | Amazon Nova Lite | GPT-4o mini |
|---|---|---|
| Provider | Amazon | OpenAI |
| Noometry Index | 31.9 | 25.5 |
| Rank | 265 | 343 |
| Context | 300K | 128K |
| Input $/M | $0.06 | $0.15 |
| Output $/M | $0.24 | $0.60 |
| Weights | Proprietary | Proprietary |

## Coding

- Amazon Nova Lite: 32.5 (#270)
- GPT-4o mini: 22.0 (#335)

| Benchmark | Amazon Nova Lite | GPT-4o mini |
|---|---|---|
| LiveBench Coding | 27.5% | 43.1% |
| LMArena Coding | 1239 | 1290 |
| Aider Polyglot | — | 3.6% |
| WeirdML | — | 11.8% |
| BigCodeBench Instruct | — | 46.1% |
| BigCodeBench Complete | — | 57.4% |
| ALE-Bench | 236.25 | — |
| HumanEval+ | — | 83.5% |
| MBPP+ | — | 72.2% |

## Agentic & Tool Use

- Amazon Nova Lite: —
- GPT-4o mini: 27.5 (#101)

| Benchmark | Amazon Nova Lite | GPT-4o mini |
|---|---|---|
| BALROG | — | 17.4% |

## Reasoning

- Amazon Nova Lite: 19.3 (#260)
- GPT-4o mini: 8.7 (#347)

| Benchmark | Amazon Nova Lite | GPT-4o mini |
|---|---|---|
| LiveBench Reasoning | 36.7% | 32.8% |
| LMArena Hard Prompts | 1220 | 1267 |
| LiveBench Data Analysis | 37.2% | 50% |
| LiveBench | 36.4% | 41.3% |
| ARC-AGI-2 | — | 0% |
| SimpleBench | — | 10.7% |
| Kagi LLM Benchmark | — | 28.8% |
| Chess Puzzles | — | 0% |
| Mystery Game Puzzles | — | 12% |
| DTBench | — | 54.4% |
| LMCA | — | 10.4% |
| Epoch Capabilities Index | — | 126.56 |
| PIQA | — | 88.7% |

## Math

- Amazon Nova Lite: 27.8 (#247)
- GPT-4o mini: 10.4 (#314)

| Benchmark | Amazon Nova Lite | GPT-4o mini |
|---|---|---|
| Omni-MATH | 23.3% | 28% |
| LiveBench Math | 36.7% | 36.3% |
| LMArena Math | 1227 | 1267 |
| FrontierMath (Tiers 1-3) | — | 0.7% |
| OTIS Mock AIME 2024-2025 | — | 6.9% |
| MATH Level 5 | — | 52.6% |
| GSM8K | — | 91.3% |

## Knowledge

- Amazon Nova Lite: 26.3 (#259)
- GPT-4o mini: 17.7 (#284)

| Benchmark | Amazon Nova Lite | GPT-4o mini |
|---|---|---|
| MMLU-Pro | 60% | 60.3% |
| GPQA (HELM) | 39.7% | 36.8% |
| LMArena Expert | 1201 | 1235 |
| MMLU | 77% | 81.8% |
| GPQA Diamond | — | 37.7% |
| Humanity's Last Exam | 3.6% | — |
| SimpleQA Verified | — | 8.3% |
| Confabulations | — | 37.2% |
| Vectara Hallucination Rate | 6.1% | — |
| BoolQ | — | 88.7% |

## Multimodal

- Amazon Nova Lite: 25.5 (#123)
- GPT-4o mini: 25.9 (#122)

| Benchmark | Amazon Nova Lite | GPT-4o mini |
|---|---|---|
| LMArena Vision | 990 | 1066 |
| Video-MME | — | 64.8% |
| GeoBench | — | 64% |
| VPCT | — | 34% |

## Multilingual

- Amazon Nova Lite: 38.0 (#232)
- GPT-4o mini: 42.0 (#199)

| Benchmark | Amazon Nova Lite | GPT-4o mini |
|---|---|---|
| LMArena Non-English | 1208 | 1266 |
| LMArena Chinese | 1225 | 1265 |
| LMArena French | 1237 | 1297 |
| LMArena German | 1229 | 1272 |
| LMArena Japanese | 1153 | 1216 |
| LMArena Korean | 1153 | 1195 |
| LMArena Russian | 1216 | 1275 |
| LMArena Spanish | 1226 | 1276 |

## Instruction Following

- Amazon Nova Lite: 59.3 (#255)
- GPT-4o mini: 61.9 (#239)

| Benchmark | Amazon Nova Lite | GPT-4o mini |
|---|---|---|
| LiveBench Instruction Following | 54.1% | 56.8% |
| IFEval | 77.6% | 78.2% |
| LMArena Instruction Following | 1205 | 1258 |

## Long Context

- Amazon Nova Lite: 37.4 (#217)
- GPT-4o mini: 39.1 (#186)

| Benchmark | Amazon Nova Lite | GPT-4o mini |
|---|---|---|
| LMArena Longer Query | 1234 | 1289 |

## Writing & Preference

- Amazon Nova Lite: 42.1 (#237)
- GPT-4o mini: 39.5 (#248)

| Benchmark | Amazon Nova Lite | GPT-4o mini |
|---|---|---|
| LMArena Text | 1229 | 1286 |
| LMArena Creative Writing | 1197 | 1268 |
| WildBench | 75% | 79.1% |
| LMArena Multi-Turn | 1199 | 1285 |
| LiveBench Language | 25.9% | 28.6% |
| Short-Story Creative Writing | — | 67.2% |
| EQ-Bench Creative Writing | — | 873 |

## FAQ

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

Amazon Nova Lite is the stronger model overall, scoring 31.9 to 25.5 on the Noometry Index.

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

Amazon Nova Lite is cheaper. It lists at $0.06 per million input tokens and $0.24 per million output tokens; GPT-4o mini lists at $0.15 and $0.60.

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

Amazon Nova Lite scores higher on coding benchmarks: 32.5 versus 22.0 in the Noometry coding category.

### Which has the bigger context window?

Amazon Nova Lite does, with 300K tokens against 128K.

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

31 benchmarks have published results for both models. Amazon Nova Lite has 34 scored results on Noometry and GPT-4o mini has 60.