# Hy3 vs Qwen2.5-VL 72B Instruct

> Hy3 is the stronger model overall, scoring 43.6 to 29.4 on the Noometry Index.

- Canonical page: https://noometry.com/compare/hy3-vs-qwen2-5-vl-72b-instruct
- Last updated: 2026-10-10
- Shared benchmarks: 0

## Summary

- The widest gap is in reasoning, where Hy3 leads 25.0 to 20.0.
- Hy3 is cheaper at $0.13 / $0.53 per million input/output tokens, against $2.80 / $8.40 for Qwen2.5-VL 72B Instruct.
- Hy3 accepts more context: 262K tokens versus 131K.

## Snapshot

| | Hy3 | Qwen2.5-VL 72B Instruct |
|---|---|---|
| Provider | Tencent | Alibaba (Qwen) |
| Noometry Index | 43.6 | 29.4 |
| Rank | 79 | 298 |
| Context | 262K | 131K |
| Input $/M | $0.13 | $2.80 |
| Output $/M | $0.53 | $8.40 |
| Weights | Open | Open |

## Coding

- Hy3: 46.6 (#60)
- Qwen2.5-VL 72B Instruct: —

| Benchmark | Hy3 | Qwen2.5-VL 72B Instruct |
|---|---|---|
| LMArena WebDev | 1509 | — |
| LMArena Coding | 1465 | — |

## Agentic & Tool Use

- Hy3: —
- Qwen2.5-VL 72B Instruct: 18.4 (#136)

| Benchmark | Hy3 | Qwen2.5-VL 72B Instruct |
|---|---|---|
| OSWorld | — | 5% |

## Reasoning

- Hy3: 25.0 (#134)
- Qwen2.5-VL 72B Instruct: 20.0 (#228)

| Benchmark | Hy3 | Qwen2.5-VL 72B Instruct |
|---|---|---|
| Kagi LLM Benchmark | — | 36% |
| NYT Connections (extended) | 41.2% | — |
| LMArena Hard Prompts | 1449 | — |

## Math

- Hy3: 37.4 (#98)
- Qwen2.5-VL 72B Instruct: —

| Benchmark | Hy3 | Qwen2.5-VL 72B Instruct |
|---|---|---|
| LMArena Math | 1475 | — |

## Knowledge

- Hy3: 40.2 (#108)
- Qwen2.5-VL 72B Instruct: —

| Benchmark | Hy3 | Qwen2.5-VL 72B Instruct |
|---|---|---|
| LMArena Expert | 1461 | — |

## Multimodal

- Hy3: —
- Qwen2.5-VL 72B Instruct: 32.9 (#92)

| Benchmark | Hy3 | Qwen2.5-VL 72B Instruct |
|---|---|---|
| LMArena Vision | — | 1107 |
| Video-MME | — | 73.5% |
| GeoBench | — | 62% |
| SpatialViz-Bench | — | 33.3% |

## Multilingual

- Hy3: 53.7 (#60)
- Qwen2.5-VL 72B Instruct: —

| Benchmark | Hy3 | Qwen2.5-VL 72B Instruct |
|---|---|---|
| LMArena Non-English | 1427 | — |
| LMArena Chinese | 1496 | — |
| LMArena French | 1461 | — |
| LMArena Korean | 1390 | — |
| LMArena Russian | 1435 | — |
| LMArena Spanish | 1448 | — |

## Instruction Following

- Hy3: 75.3 (#68)
- Qwen2.5-VL 72B Instruct: —

| Benchmark | Hy3 | Qwen2.5-VL 72B Instruct |
|---|---|---|
| LMArena Instruction Following | 1425 | — |

## Long Context

- Hy3: 44.1 (#70)
- Qwen2.5-VL 72B Instruct: —

| Benchmark | Hy3 | Qwen2.5-VL 72B Instruct |
|---|---|---|
| LMArena Longer Query | 1443 | — |

## Writing & Preference

- Hy3: 62.5 (#75)
- Qwen2.5-VL 72B Instruct: —

| Benchmark | Hy3 | Qwen2.5-VL 72B Instruct |
|---|---|---|
| LMArena Text | 1442 | — |
| LMArena Creative Writing | 1399 | — |
| LMArena Multi-Turn | 1436 | — |

## FAQ

### Is Hy3 better than Qwen2.5-VL 72B Instruct?

Hy3 is the stronger model overall, scoring 43.6 to 29.4 on the Noometry Index.

### Which is cheaper, Hy3 or Qwen2.5-VL 72B Instruct?

Hy3 is cheaper. It lists at $0.13 per million input tokens and $0.53 per million output tokens; Qwen2.5-VL 72B Instruct lists at $2.80 and $8.40.

### Which has the bigger context window?

Hy3 does, with 262K tokens against 131K.

### How many benchmarks do Hy3 and Qwen2.5-VL 72B Instruct share?

0 benchmarks have published results for both models. Hy3 has 17 scored results on Noometry and Qwen2.5-VL 72B Instruct has 6.