# Amazon Nova Micro vs Llama 3-70B

> Amazon Nova Micro is the stronger model overall, scoring 29.9 to 28.5 on the Noometry Index.

- Canonical page: https://noometry.com/compare/amazon-nova-micro-vs-llama-3-70b
- Last updated: 2026-10-10
- Shared benchmarks: 18

## Summary

- They share 18 benchmarks with published results for both. Amazon Nova Micro scores higher in 5 categories and Llama 3-70B in 4 categories; 7 gaps are clear of the uncertainty.
- The widest gap is in math, where Amazon Nova Micro leads 24.9 to 12.0.
- Llama 3-70B has downloadable open weights; the other is API-only.

## Snapshot

| | Amazon Nova Micro | Llama 3-70B |
|---|---|---|
| Provider | Amazon | Meta |
| Noometry Index | 29.9 | 28.5 |
| Rank | 290 | 317 |
| Context | 128K | — |
| Input $/M | $0.035 | — |
| Output $/M | $0.14 | — |
| Weights | Proprietary | Open |

## Coding

- Amazon Nova Micro: 30.0 (#289)
- Llama 3-70B: 35.3 (#215)

| Benchmark | Amazon Nova Micro | Llama 3-70B |
|---|---|---|
| LMArena Coding | 1218 | 1206 |
| BigCodeBench Instruct | — | 43.6% |
| LiveBench Coding | 20.2% | — |
| BigCodeBench Complete | — | 54.5% |
| HumanEval+ | — | 72% |
| MBPP+ | — | 69% |

## Agentic & Tool Use

- Amazon Nova Micro: 21.9 (#125)
- Llama 3-70B: 20.9 (#132)

| Benchmark | Amazon Nova Micro | Llama 3-70B |
|---|---|---|
| Berkeley Function Calling Leaderboard | 22.3% | — |
| Cybench | — | 5% |

## Reasoning

- Amazon Nova Micro: 16.7 (#289)
- Llama 3-70B: 17.4 (#284)

| Benchmark | Amazon Nova Micro | Llama 3-70B |
|---|---|---|
| LMArena Hard Prompts | 1191 | 1195 |
| Kagi LLM Benchmark | — | 35.1% |
| LiveBench Reasoning | 25.1% | — |
| DTBench | — | 54.2% |
| LiveBench Data Analysis | 34% | — |
| Epoch Capabilities Index | — | 122.91 |
| ForecastBench | — | 57.1 |
| LiveBench | 29.6% | — |
| WinoGrande | — | 83.5% |

## Math

- Amazon Nova Micro: 24.9 (#252)
- Llama 3-70B: 12.0 (#299)

| Benchmark | Amazon Nova Micro | Llama 3-70B |
|---|---|---|
| LMArena Math | 1206 | 1218 |
| OTIS Mock AIME 2024-2025 | — | 4.3% |
| Omni-MATH | 21.4% | — |
| LiveBench Math | 34.5% | — |
| MATH Level 5 | — | 22.6% |

## Knowledge

- Amazon Nova Micro: 29.2 (#231)
- Llama 3-70B: 20.8 (#272)

| Benchmark | Amazon Nova Micro | Llama 3-70B |
|---|---|---|
| LMArena Expert | 1184 | 1149 |
| MMLU | 70.8% | 79.3% |
| GPQA Diamond | — | 40.6% |
| MMLU-Pro | 51.1% | — |
| Vectara Hallucination Rate | 5.5% | — |
| GPQA (HELM) | 38.3% | — |

## Multilingual

- Amazon Nova Micro: 36.7 (#234)
- Llama 3-70B: 33.8 (#246)

| Benchmark | Amazon Nova Micro | Llama 3-70B |
|---|---|---|
| LMArena Non-English | 1186 | 1142 |
| LMArena Chinese | 1209 | 1114 |
| LMArena French | 1237 | 1231 |
| LMArena German | 1192 | 1169 |
| LMArena Japanese | 1154 | 1017 |
| LMArena Korean | 1150 | 1017 |
| LMArena Russian | 1185 | 1159 |
| LMArena Spanish | 1225 | 1241 |

## Instruction Following

- Amazon Nova Micro: 56.6 (#267)
- Llama 3-70B: 62.8 (#233)

| Benchmark | Amazon Nova Micro | Llama 3-70B |
|---|---|---|
| LMArena Instruction Following | 1174 | 1194 |
| LiveBench Instruction Following | 48% | — |
| IFEval | 76% | — |

## Long Context

- Amazon Nova Micro: 36.5 (#224)
- Llama 3-70B: 35.6 (#235)

| Benchmark | Amazon Nova Micro | Llama 3-70B |
|---|---|---|
| LMArena Longer Query | 1206 | 1174 |

## Writing & Preference

- Amazon Nova Micro: 39.7 (#242)
- Llama 3-70B: 43.0 (#225)

| Benchmark | Amazon Nova Micro | Llama 3-70B |
|---|---|---|
| LMArena Text | 1208 | 1221 |
| LMArena Creative Writing | 1172 | 1210 |
| LMArena Multi-Turn | 1178 | 1223 |
| WildBench | 74.3% | — |
| LiveBench Language | 15.8% | — |

## FAQ

### Is Amazon Nova Micro better than Llama 3-70B?

Amazon Nova Micro is the stronger model overall, scoring 29.9 to 28.5 on the Noometry Index.

### Is Amazon Nova Micro or Llama 3-70B better for coding?

Llama 3-70B scores higher on coding benchmarks: 35.3 versus 30.0 in the Noometry coding category.

### How many benchmarks do Amazon Nova Micro and Llama 3-70B share?

18 benchmarks have published results for both models. Amazon Nova Micro has 32 scored results on Noometry and Llama 3-70B has 31.