# Amazon Nova Micro vs GPT-6 Astra

> GPT-6 Astra is the stronger model overall, scoring 70.8 to 30.4 on the Noometry Index. Amazon Nova Micro costs 327× less per token, which makes it the better buy when GPT-6 Astra's lead doesn't matter for your workload.

- Canonical page: https://noometry.com/compare/amazon-nova-micro-vs-gpt-6-astra
- Last updated: 2026-10-10
- Shared benchmarks: 18

## Summary

- They share 18 benchmarks with published results for both. Amazon Nova Micro scores higher in 0 categories and GPT-6 Astra in 9 categories; 9 gaps are clear of the uncertainty.
- The widest gap is in reasoning, where GPT-6 Astra leads 85.1 to 17.4.
- Amazon Nova Micro is cheaper at $0.035 / $0.14 per million input/output tokens, against $10 / $50 for GPT-6 Astra.
- GPT-6 Astra accepts more context: 1.05M tokens versus 128K.

## Snapshot

| | Amazon Nova Micro | GPT-6 Astra |
|---|---|---|
| Provider | Amazon | OpenAI |
| Noometry Index | 30.4 | 70.8 |
| Rank | 294 | 1 |
| Context | 128K | 1.05M |
| Input $/M | $0.035 | $10 |
| Output $/M | $0.14 | $50 |
| Weights | Proprietary | Proprietary |

## Coding

- Amazon Nova Micro: 30.5 (#295)
- GPT-6 Astra: 73.7 (#2)

| Benchmark | Amazon Nova Micro | GPT-6 Astra |
|---|---|---|
| LMArena Coding | 1218 | 1487 |
| DeepSWE | — | 74.1% |
| FrontierCode | — | 53.3% |
| LMArena WebDev | — | 1786 |
| FrontierSWE | — | 65.5% |
| SciCode | — | 56.5% |
| GSO | — | 79.4% |
| WeirdML | — | 93.6% |
| LiveBench Coding | 20.2% | — |
| MirrorCode | — | 46.7% |
| ALE-Bench | — | 2,951 |

## Agentic & Tool Use

- Amazon Nova Micro: 22.1 (#132)
- GPT-6 Astra: 52.9 (#3)

| Benchmark | Amazon Nova Micro | GPT-6 Astra |
|---|---|---|
| APEX-Agents | — | 64.7% |
| Berkeley Function Calling Leaderboard | 22.3% | — |
| Remote Labor Index | — | 20.8% |
| BALROG | — | 68.3% |
| GDP.pdf | — | 34.2% |
| Vending-Bench 2 | — | 15,515 |

## Reasoning

- Amazon Nova Micro: 17.4 (#294)
- GPT-6 Astra: 85.1 (#1)

| Benchmark | Amazon Nova Micro | GPT-6 Astra |
|---|---|---|
| LMArena Hard Prompts | 1191 | 1462 |
| ARC-AGI-2 | — | 95% |
| NYT Connections (extended) | — | 98.1% |
| ARC-AGI-1 | — | 98.5% |
| CritPt | — | 31.7% |
| Chess Puzzles | — | 72% |
| EBR-Bench | — | 76.2% |
| LiveBench Reasoning | 25.1% | — |
| Mystery Game Puzzles | — | 84% |
| DTBench | — | 97.3% |
| LiveBench Data Analysis | 34% | — |
| LMCA | — | 64.4% |
| Bench to the Future 3 | — | 0.14 |
| Epoch Capabilities Index | — | 166.45 |
| LiveBench | 29.6% | — |

## Math

- Amazon Nova Micro: 26.9 (#254)
- GPT-6 Astra: 93.5 (#2)

| Benchmark | Amazon Nova Micro | GPT-6 Astra |
|---|---|---|
| LMArena Math | 1206 | 1465 |
| FrontierMath (Tiers 1-3) | — | 93.7% |
| FrontierMath Tier 4 | — | 97.6% |
| OTIS Mock AIME 2024-2025 | — | 100% |
| ProofBench | — | 99% |
| Omni-MATH | 21.4% | — |
| LiveBench Math | 34.5% | — |
| FrontierMath Erdős | — | 2.9% |

## Knowledge

- Amazon Nova Micro: 29.6 (#237)
- GPT-6 Astra: 75.3 (#1)

| Benchmark | Amazon Nova Micro | GPT-6 Astra |
|---|---|---|
| Vectara Hallucination Rate | 5.5% | 8.7% |
| LMArena Expert | 1184 | 1483 |
| GPQA Diamond | — | 95.8% |
| Humanity's Last Exam | — | 54.8% |
| SimpleQA Verified | — | 75.6% |
| MMLU-Pro | 51.1% | — |
| GPQA (HELM) | 38.3% | — |
| MMLU | 70.8% | — |

## Multimodal

- Amazon Nova Micro: —
- GPT-6 Astra: 55.0 (#3)

| Benchmark | Amazon Nova Micro | GPT-6 Astra |
|---|---|---|
| LMArena Vision | — | 1281 |
| Blueprint-Bench 2 | — | 49.7% |
| Furniture Assembly | — | 80% |
| LMArena Document | — | 1468 |

## Multilingual

- Amazon Nova Micro: 36.5 (#239)
- GPT-6 Astra: 53.7 (#61)

| Benchmark | Amazon Nova Micro | GPT-6 Astra |
|---|---|---|
| LMArena Non-English | 1186 | 1430 |
| LMArena Chinese | 1209 | 1484 |
| LMArena French | 1238 | 1456 |
| LMArena German | 1192 | 1440 |
| LMArena Japanese | 1154 | 1379 |
| LMArena Korean | 1150 | 1426 |
| LMArena Russian | 1185 | 1436 |
| LMArena Spanish | 1225 | 1407 |

## Instruction Following

- Amazon Nova Micro: 56.3 (#272)
- GPT-6 Astra: 76.3 (#44)

| Benchmark | Amazon Nova Micro | GPT-6 Astra |
|---|---|---|
| LMArena Instruction Following | 1174 | 1450 |
| LiveBench Instruction Following | 48% | — |
| IFEval | 76% | — |

## Long Context

- Amazon Nova Micro: 36.5 (#229)
- GPT-6 Astra: 44.5 (#62)

| Benchmark | Amazon Nova Micro | GPT-6 Astra |
|---|---|---|
| LMArena Longer Query | 1205 | 1456 |

## Writing & Preference

- Amazon Nova Micro: 39.5 (#247)
- GPT-6 Astra: 75.3 (#7)

| Benchmark | Amazon Nova Micro | GPT-6 Astra |
|---|---|---|
| LMArena Text | 1208 | 1441 |
| LMArena Creative Writing | 1172 | 1418 |
| LMArena Multi-Turn | 1178 | 1448 |
| EQ-Bench Creative Writing | — | 2173 |
| WildBench | 74.3% | — |
| LiveBench Language | 15.8% | — |

## FAQ

### Is Amazon Nova Micro better than GPT-6 Astra?

GPT-6 Astra is the stronger model overall, scoring 70.8 to 30.4 on the Noometry Index. Amazon Nova Micro costs 327× less per token, which makes it the better buy when GPT-6 Astra's lead doesn't matter for your workload.

### Which is cheaper, Amazon Nova Micro or GPT-6 Astra?

Amazon Nova Micro is cheaper. It lists at $0.035 per million input tokens and $0.14 per million output tokens; GPT-6 Astra lists at $10 and $50.

### Is Amazon Nova Micro or GPT-6 Astra better for coding?

GPT-6 Astra scores higher on coding benchmarks: 73.7 versus 30.5 in the Noometry coding category.

### Which has the bigger context window?

GPT-6 Astra does, with 1.05M tokens against 128K.

### How many benchmarks do Amazon Nova Micro and GPT-6 Astra share?

18 benchmarks have published results for both models. Amazon Nova Micro has 32 scored results on Noometry and GPT-6 Astra has 56.