# Amazon Nova Experimental Chat 10 20 vs GPT-4

> Amazon Nova Experimental Chat 10 20 is the stronger model overall, scoring 41.5 to 28.9 on the Noometry Index.

- Canonical page: https://noometry.com/compare/amazon-nova-experimental-chat-10-20-vs-gpt-4
- Last updated: 2026-10-10
- Shared benchmarks: 17

## Summary

- They share 17 benchmarks with published results for both. Amazon Nova Experimental Chat 10 20 scores higher in 8 categories and GPT-4 in 0 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in math, where Amazon Nova Experimental Chat 10 20 leads 36.2 to 10.1.

## Snapshot

| | Amazon Nova Experimental Chat 10 20 | GPT-4 |
|---|---|---|
| Provider | Amazon | OpenAI |
| Noometry Index | 41.5 | 28.9 |
| Rank | 122 | 308 |
| Context | — | 8K |
| Input $/M | — | $30 |
| Output $/M | — | $60 |
| Weights | Proprietary | Proprietary |

## Coding

- Amazon Nova Experimental Chat 10 20: 41.0 (#123)
- GPT-4: 31.2 (#277)

| Benchmark | Amazon Nova Experimental Chat 10 20 | GPT-4 |
|---|---|---|
| LMArena Coding | 1413 | 1254 |
| WeirdML | — | 12.4% |
| BigCodeBench Instruct | — | 46% |
| BigCodeBench Complete | — | 57.2% |
| HumanEval+ | — | 79.3% |

## Agentic & Tool Use

- Amazon Nova Experimental Chat 10 20: —
- GPT-4: —

| Benchmark | Amazon Nova Experimental Chat 10 20 | GPT-4 |
|---|---|---|
| METR Time Horizons | — | 36.1% |

## Reasoning

- Amazon Nova Experimental Chat 10 20: 27.3 (#110)
- GPT-4: 17.3 (#285)

| Benchmark | Amazon Nova Experimental Chat 10 20 | GPT-4 |
|---|---|---|
| LMArena Hard Prompts | 1396 | 1241 |
| Chess Puzzles | — | 4% |
| Mystery Game Puzzles | — | 12% |
| DTBench | — | 62.7% |
| LMCA | — | 17.1% |
| BIG-Bench Hard | — | 75.1% |
| Epoch Capabilities Index | — | 125.88 |
| ForecastBench | — | 57.8 |
| HellaSwag | — | 95.3% |
| WinoGrande | — | 87.5% |

## Math

- Amazon Nova Experimental Chat 10 20: 36.2 (#121)
- GPT-4: 10.1 (#304)

| Benchmark | Amazon Nova Experimental Chat 10 20 | GPT-4 |
|---|---|---|
| LMArena Math | 1425 | 1269 |
| OTIS Mock AIME 2024-2025 | — | 1.1% |
| MATH Level 5 | — | 23% |
| GSM8K | — | 92% |

## Knowledge

- Amazon Nova Experimental Chat 10 20: 37.8 (#142)
- GPT-4: 18.4 (#277)

| Benchmark | Amazon Nova Experimental Chat 10 20 | GPT-4 |
|---|---|---|
| LMArena Expert | 1385 | 1211 |
| GPQA Diamond | — | 35.7% |
| MMLU | — | 86.4% |
| TriviaQA | — | 84.8% |

## Multilingual

- Amazon Nova Experimental Chat 10 20: 49.7 (#127)
- GPT-4: 40.8 (#210)

| Benchmark | Amazon Nova Experimental Chat 10 20 | GPT-4 |
|---|---|---|
| LMArena Non-English | 1372 | 1246 |
| LMArena Chinese | 1405 | 1242 |
| LMArena French | 1426 | 1282 |
| LMArena German | 1391 | 1251 |
| LMArena Japanese | 1355 | 1209 |
| LMArena Korean | 1327 | 1184 |
| LMArena Russian | 1371 | 1251 |
| LMArena Spanish | 1381 | 1261 |

## Instruction Following

- Amazon Nova Experimental Chat 10 20: 72.3 (#136)
- GPT-4: 65.5 (#217)

| Benchmark | Amazon Nova Experimental Chat 10 20 | GPT-4 |
|---|---|---|
| LMArena Instruction Following | 1365 | 1241 |

## Long Context

- Amazon Nova Experimental Chat 10 20: 41.7 (#128)
- GPT-4: 37.7 (#207)

| Benchmark | Amazon Nova Experimental Chat 10 20 | GPT-4 |
|---|---|---|
| LMArena Longer Query | 1370 | 1244 |

## Writing & Preference

- Amazon Nova Experimental Chat 10 20: 56.9 (#131)
- GPT-4: 35.0 (#263)

| Benchmark | Amazon Nova Experimental Chat 10 20 | GPT-4 |
|---|---|---|
| LMArena Text | 1394 | 1263 |
| LMArena Creative Writing | 1318 | 1244 |
| LMArena Multi-Turn | 1364 | 1257 |
| EQ-Bench Creative Writing | — | 752 |

## FAQ

### Is Amazon Nova Experimental Chat 10 20 better than GPT-4?

Amazon Nova Experimental Chat 10 20 is the stronger model overall, scoring 41.5 to 28.9 on the Noometry Index.

### Is Amazon Nova Experimental Chat 10 20 or GPT-4 better for coding?

Amazon Nova Experimental Chat 10 20 scores higher on coding benchmarks: 41.0 versus 31.2 in the Noometry coding category.

### How many benchmarks do Amazon Nova Experimental Chat 10 20 and GPT-4 share?

17 benchmarks have published results for both models. Amazon Nova Experimental Chat 10 20 has 17 scored results on Noometry and GPT-4 has 38.