Model comparison
Hy3 vs Qwen3.5-Flash
Hy3 is the stronger model overall, scoring 44.2 to 42.5 on the Noometry Index.
Last verified . 18 shared benchmarks.
Tencent
Alibaba (Qwen)Summary
- They share 18 benchmarks with published results for both. Hy3 scores higher in 6 categories and Qwen3.5-Flash in 2 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in coding, where Hy3 leads 46.8 to 34.2.
- Hy3 is cheaper at $0.0825 / $0.33 per million input/output tokens, against $0.10 / $0.40 for Qwen3.5-Flash.
- Qwen3.5-Flash accepts more context: 1M tokens versus 262K.
- Hy3 has downloadable open weights; the other is API-only.
Side by side
| Hy3 | Qwen3.5-Flash | |
|---|---|---|
| Provider | Tencent | Alibaba (Qwen) |
| Noometry Index | 44.2 | 42.5 |
| Released | 2026-07-06 | 2026-02-23 |
| Weights | Open | Proprietary |
| Context window | 262K | 1M |
| Max output | 128K | 66K |
| Input $ / M tokens | $0.0825 | $0.10 |
| Output $ / M tokens | $0.33 | $0.40 |
| Results tracked | 19 | 32 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Hy3 leads
Hy3: 46.8 (#63), Qwen3.5-Flash: 34.2 (#242)
| Benchmark | Hy3 | Qwen3.5-Flash |
|---|---|---|
| LMArena WebDev | 1508 | 1244 |
| LMArena Coding | 1464 | 1412 |
| ALE-Bench | — | 221.8 |
Agentic & Tool Use Not comparable
Hy3: —, Qwen3.5-Flash: —
| Benchmark | Hy3 | Qwen3.5-Flash |
|---|---|---|
| Vending-Bench 2 | — | 462.69 |
Reasoning Qwen3.5-Flash leads
Hy3: 26.1 (#136), Qwen3.5-Flash: 33.7 (#72)
| Benchmark | Hy3 | Qwen3.5-Flash |
|---|---|---|
| LMArena Hard Prompts | 1447 | 1403 |
| NYT Connections (extended) | 41.2% | — |
| Chess Puzzles | — | 21% |
| Mystery Game Puzzles | — | 20% |
| DTBench | — | 82.9% |
| LMCA | — | 29.1% |
| Epoch Capabilities Index | — | 143.98 |
Math Hy3 leads
Hy3: 40.1 (#93), Qwen3.5-Flash: 37.4 (#158)
| Benchmark | Hy3 | Qwen3.5-Flash |
|---|---|---|
| LMArena Math | 1475 | 1407 |
| FrontierMath (Tiers 1-3) | — | 18.2% |
| OTIS Mock AIME 2024-2025 | — | 84.4% |
| FrontierMath (Feb 2025 set) | — | 6.2% |
| FrontierMath Tier 4 (v1) | — | 0% |
Knowledge Qwen3.5-Flash leads
Hy3: 40.8 (#114), Qwen3.5-Flash: 43.2 (#93)
| Benchmark | Hy3 | Qwen3.5-Flash |
|---|---|---|
| LMArena Expert | 1460 | 1407 |
| GPQA Diamond | — | 82.3% |
| SimpleQA Verified | — | 20.3% |
| Vectara Hallucination Rate | — | 10.5% |
Multilingual Hy3 leads
Hy3: 53.5 (#65), Qwen3.5-Flash: 50.5 (#121)
| Benchmark | Hy3 | Qwen3.5-Flash |
|---|---|---|
| LMArena Non-English | 1426 | 1385 |
| LMArena Chinese | 1493 | 1446 |
| LMArena French | 1461 | 1412 |
| LMArena German | 1439 | 1390 |
| LMArena Japanese | 1392 | 1368 |
| LMArena Korean | 1395 | 1344 |
| LMArena Russian | 1432 | 1379 |
| LMArena Spanish | 1456 | 1400 |
Instruction Following Hy3 leads
Hy3: 75.1 (#70), Qwen3.5-Flash: 72.6 (#139)
| Benchmark | Hy3 | Qwen3.5-Flash |
|---|---|---|
| LMArena Instruction Following | 1426 | 1374 |
Long Context Hy3 leads
Hy3: 44.1 (#75), Qwen3.5-Flash: 42.4 (#124)
| Benchmark | Hy3 | Qwen3.5-Flash |
|---|---|---|
| LMArena Longer Query | 1442 | 1392 |
Writing & Preference Hy3 leads
Hy3: 62.2 (#81), Qwen3.5-Flash: 57.9 (#122)
| Benchmark | Hy3 | Qwen3.5-Flash |
|---|---|---|
| LMArena Text | 1439 | 1397 |
| LMArena Creative Writing | 1402 | 1343 |
| LMArena Multi-Turn | 1436 | 1393 |
Frequently asked questions
Is Hy3 better than Qwen3.5-Flash?
Hy3 is the stronger model overall, scoring 44.2 to 42.5 on the Noometry Index.
Which is cheaper, Hy3 or Qwen3.5-Flash?
Hy3 is cheaper. It lists at $0.0825 per million input tokens and $0.33 per million output tokens; Qwen3.5-Flash lists at $0.10 and $0.40.
Is Hy3 or Qwen3.5-Flash better for coding?
Hy3 scores higher on coding benchmarks: 46.8 versus 34.2 in the Noometry coding category.
Which has the bigger context window?
Qwen3.5-Flash does, with 1M tokens against 262K.
How many benchmarks do Hy3 and Qwen3.5-Flash share?
18 benchmarks have published results for both models. Hy3 has 19 scored results on Noometry and Qwen3.5-Flash has 32.