# Hy3 vs Qwen3-30B-A3B

> Hy3 is the stronger model overall, scoring 44.2 to 38.9 on the Noometry Index.

- Canonical page: https://noometry.com/compare/hy3-vs-qwen3-30b-a3b
- Last updated: 2026-10-10
- Shared benchmarks: 17

## Summary

- They share 17 benchmarks with published results for both. Hy3 scores higher in 7 categories and Qwen3-30B-A3B in 1 category; 7 gaps are clear of the uncertainty.
- The widest gap is in long context, where Hy3 leads 44.1 to 31.0.
- Hy3 is cheaper at $0.0825 / $0.33 per million input/output tokens, against $0.12 / $0.50 for Qwen3-30B-A3B.
- Hy3 accepts more context: 262K tokens versus 41K.

## Snapshot

| | Hy3 | Qwen3-30B-A3B |
|---|---|---|
| Provider | Tencent | Alibaba (Qwen) |
| Noometry Index | 44.2 | 38.9 |
| Rank | 79 | 179 |
| Context | 262K | 41K |
| Input $/M | $0.0825 | $0.12 |
| Output $/M | $0.33 | $0.50 |
| Weights | Open | Open |

## Coding

- Hy3: 46.8 (#63)
- Qwen3-30B-A3B: 37.5 (#194)

| Benchmark | Hy3 | Qwen3-30B-A3B |
|---|---|---|
| LMArena Coding | 1464 | 1416 |
| LMArena WebDev | 1508 | — |
| SciCode | — | 33.3% |
| WeirdML | — | 29.8% |

## Agentic & Tool Use

- Hy3: —
- Qwen3-30B-A3B: 29.8 (#82)

| Benchmark | Hy3 | Qwen3-30B-A3B |
|---|---|---|
| Berkeley Function Calling Leaderboard | — | 41.4% |

## Reasoning

- Hy3: 26.1 (#136)
- Qwen3-30B-A3B: 22.2 (#204)

| Benchmark | Hy3 | Qwen3-30B-A3B |
|---|---|---|
| LMArena Hard Prompts | 1447 | 1398 |
| Kagi LLM Benchmark | — | 54.9% |
| NYT Connections (extended) | 41.2% | — |
| CritPt | — | 0.3% |
| Chess Puzzles | — | 8% |
| DTBench | — | 69.3% |
| LMCA | — | 22.4% |
| Epoch Capabilities Index | — | 139.63 |

## Math

- Hy3: 40.1 (#93)
- Qwen3-30B-A3B: 37.4 (#157)

| Benchmark | Hy3 | Qwen3-30B-A3B |
|---|---|---|
| LMArena Math | 1475 | 1394 |
| MathArena Final-Answer Competitions | — | 47.8% |
| OTIS Mock AIME 2024-2025 | — | 70.3% |

## Knowledge

- Hy3: 40.8 (#114)
- Qwen3-30B-A3B: 41.8 (#105)

| Benchmark | Hy3 | Qwen3-30B-A3B |
|---|---|---|
| LMArena Expert | 1460 | 1396 |
| GPQA Diamond | — | 70.1% |
| Confabulations | — | 12.3% |

## Multilingual

- Hy3: 53.5 (#65)
- Qwen3-30B-A3B: 49.5 (#132)

| Benchmark | Hy3 | Qwen3-30B-A3B |
|---|---|---|
| LMArena Non-English | 1426 | 1372 |
| LMArena Chinese | 1493 | 1433 |
| LMArena French | 1461 | 1418 |
| LMArena German | 1439 | 1380 |
| LMArena Japanese | 1392 | 1337 |
| LMArena Korean | 1395 | 1331 |
| LMArena Russian | 1432 | 1370 |
| LMArena Spanish | 1456 | 1404 |

## Instruction Following

- Hy3: 75.1 (#70)
- Qwen3-30B-A3B: 72.0 (#142)

| Benchmark | Hy3 | Qwen3-30B-A3B |
|---|---|---|
| LMArena Instruction Following | 1426 | 1363 |

## Long Context

- Hy3: 44.1 (#75)
- Qwen3-30B-A3B: 31.0 (#283)

| Benchmark | Hy3 | Qwen3-30B-A3B |
|---|---|---|
| LMArena Longer Query | 1442 | 1379 |
| Fiction.LiveBench | — | 40.6% |

## Writing & Preference

- Hy3: 62.2 (#81)
- Qwen3-30B-A3B: 55.6 (#143)

| Benchmark | Hy3 | Qwen3-30B-A3B |
|---|---|---|
| LMArena Text | 1439 | 1384 |
| LMArena Creative Writing | 1402 | 1317 |
| LMArena Multi-Turn | 1436 | 1378 |
| Short-Story Creative Writing | — | 75.3% |

## FAQ

### Is Hy3 better than Qwen3-30B-A3B?

Hy3 is the stronger model overall, scoring 44.2 to 38.9 on the Noometry Index.

### Which is cheaper, Hy3 or Qwen3-30B-A3B?

Hy3 is cheaper. It lists at $0.0825 per million input tokens and $0.33 per million output tokens; Qwen3-30B-A3B lists at $0.12 and $0.50.

### Is Hy3 or Qwen3-30B-A3B better for coding?

Hy3 scores higher on coding benchmarks: 46.8 versus 37.5 in the Noometry coding category.

### Which has the bigger context window?

Hy3 does, with 262K tokens against 41K.

### How many benchmarks do Hy3 and Qwen3-30B-A3B share?

17 benchmarks have published results for both models. Hy3 has 19 scored results on Noometry and Qwen3-30B-A3B has 32.