# GPT-6 Astra vs Hy3

> GPT-6 Astra is the stronger model overall, scoring 70.8 to 44.2 on the Noometry Index. Hy3 costs 139× 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/gpt-6-astra-vs-hy3
- Last updated: 2026-10-10
- Shared benchmarks: 19

## Summary

- They share 19 benchmarks with published results for both. GPT-6 Astra scores higher in 8 categories and Hy3 in 0 categories; 6 gaps are clear of the uncertainty.
- The widest gap is in reasoning, where GPT-6 Astra leads 85.1 to 26.1.
- The biggest single-benchmark swing is NYT Connections (extended): 98.1% for GPT-6 Astra and 41.2% for Hy3.
- Hy3 is cheaper at $0.0825 / $0.33 per million input/output tokens, against $10 / $50 for GPT-6 Astra.
- GPT-6 Astra accepts more context: 1.05M tokens versus 262K.
- Hy3 has downloadable open weights; the other is API-only.

## Snapshot

| | GPT-6 Astra | Hy3 |
|---|---|---|
| Provider | OpenAI | Tencent |
| Noometry Index | 70.8 | 44.2 |
| Rank | 1 | 79 |
| Context | 1.05M | 262K |
| Input $/M | $10 | $0.0825 |
| Output $/M | $50 | $0.33 |
| Weights | Proprietary | Open |

## Coding

- GPT-6 Astra: 73.7 (#2)
- Hy3: 46.8 (#63)

| Benchmark | GPT-6 Astra | Hy3 |
|---|---|---|
| LMArena WebDev | 1786 | 1508 |
| LMArena Coding | 1487 | 1464 |
| DeepSWE | 74.1% | — |
| FrontierCode | 53.3% | — |
| FrontierSWE | 65.5% | — |
| SciCode | 56.5% | — |
| GSO | 79.4% | — |
| WeirdML | 93.6% | — |
| MirrorCode | 46.7% | — |
| ALE-Bench | 2,951 | — |

## Agentic & Tool Use

- GPT-6 Astra: 52.9 (#3)
- Hy3: —

| Benchmark | GPT-6 Astra | Hy3 |
|---|---|---|
| APEX-Agents | 64.7% | — |
| Remote Labor Index | 20.8% | — |
| BALROG | 68.3% | — |
| GDP.pdf | 34.2% | — |
| Vending-Bench 2 | 15,515 | — |

## Reasoning

- GPT-6 Astra: 85.1 (#1)
- Hy3: 26.1 (#136)

| Benchmark | GPT-6 Astra | Hy3 |
|---|---|---|
| NYT Connections (extended) | 98.1% | 41.2% |
| LMArena Hard Prompts | 1462 | 1447 |
| ARC-AGI-2 | 95% | — |
| ARC-AGI-1 | 98.5% | — |
| CritPt | 31.7% | — |
| Chess Puzzles | 72% | — |
| EBR-Bench | 76.2% | — |
| Mystery Game Puzzles | 84% | — |
| DTBench | 97.3% | — |
| LMCA | 64.4% | — |
| Bench to the Future 3 | 0.14 | — |
| Epoch Capabilities Index | 166.45 | — |

## Math

- GPT-6 Astra: 93.5 (#2)
- Hy3: 40.1 (#93)

| Benchmark | GPT-6 Astra | Hy3 |
|---|---|---|
| LMArena Math | 1465 | 1475 |
| FrontierMath (Tiers 1-3) | 93.7% | — |
| FrontierMath Tier 4 | 97.6% | — |
| OTIS Mock AIME 2024-2025 | 100% | — |
| ProofBench | 99% | — |
| FrontierMath Erdős | 2.9% | — |

## Knowledge

- GPT-6 Astra: 75.3 (#1)
- Hy3: 40.8 (#114)

| Benchmark | GPT-6 Astra | Hy3 |
|---|---|---|
| LMArena Expert | 1483 | 1460 |
| GPQA Diamond | 95.8% | — |
| Humanity's Last Exam | 54.8% | — |
| SimpleQA Verified | 75.6% | — |
| Vectara Hallucination Rate | 8.7% | — |

## Multimodal

- GPT-6 Astra: 55.0 (#3)
- Hy3: —

| Benchmark | GPT-6 Astra | Hy3 |
|---|---|---|
| LMArena Vision | 1281 | — |
| Blueprint-Bench 2 | 49.7% | — |
| Furniture Assembly | 80% | — |
| LMArena Document | 1468 | — |

## Multilingual

- GPT-6 Astra: 53.7 (#61)
- Hy3: 53.5 (#65)

| Benchmark | GPT-6 Astra | Hy3 |
|---|---|---|
| LMArena Non-English | 1430 | 1426 |
| LMArena Chinese | 1484 | 1493 |
| LMArena French | 1456 | 1461 |
| LMArena German | 1440 | 1439 |
| LMArena Japanese | 1379 | 1392 |
| LMArena Korean | 1426 | 1395 |
| LMArena Russian | 1436 | 1432 |
| LMArena Spanish | 1407 | 1456 |

## Instruction Following

- GPT-6 Astra: 76.3 (#44)
- Hy3: 75.1 (#70)

| Benchmark | GPT-6 Astra | Hy3 |
|---|---|---|
| LMArena Instruction Following | 1450 | 1426 |

## Long Context

- GPT-6 Astra: 44.5 (#62)
- Hy3: 44.1 (#75)

| Benchmark | GPT-6 Astra | Hy3 |
|---|---|---|
| LMArena Longer Query | 1456 | 1442 |

## Writing & Preference

- GPT-6 Astra: 75.3 (#7)
- Hy3: 62.2 (#81)

| Benchmark | GPT-6 Astra | Hy3 |
|---|---|---|
| LMArena Text | 1441 | 1439 |
| LMArena Creative Writing | 1418 | 1402 |
| LMArena Multi-Turn | 1448 | 1436 |
| EQ-Bench Creative Writing | 2173 | — |

## FAQ

### Is GPT-6 Astra better than Hy3?

GPT-6 Astra is the stronger model overall, scoring 70.8 to 44.2 on the Noometry Index. Hy3 costs 139× less per token, which makes it the better buy when GPT-6 Astra's lead doesn't matter for your workload.

### Which is cheaper, GPT-6 Astra or Hy3?

Hy3 is cheaper. It lists at $0.0825 per million input tokens and $0.33 per million output tokens; GPT-6 Astra lists at $10 and $50.

### Is GPT-6 Astra or Hy3 better for coding?

GPT-6 Astra scores higher on coding benchmarks: 73.7 versus 46.8 in the Noometry coding category.

### Which has the bigger context window?

GPT-6 Astra does, with 1.05M tokens against 262K.

### How many benchmarks do GPT-6 Astra and Hy3 share?

19 benchmarks have published results for both models. GPT-6 Astra has 56 scored results on Noometry and Hy3 has 19.