Model comparison
DeepSeek-V3.2-Speciale vs Hy3
Hy3 is the stronger model overall, scoring 43.6 to 39.1 on the Noometry Index.
Last verified . 0 shared benchmarks.
DeepSeek
TencentSummary
- The widest gap is in writing & preference, where Hy3 leads 62.5 to 46.1.
- Hy3 is cheaper at $0.13 / $0.53 per million input/output tokens, against $0.58 / $1.68 for DeepSeek-V3.2-Speciale.
- Hy3 accepts more context: 262K tokens versus 128K.
Side by side
| DeepSeek-V3.2-Speciale | Hy3 | |
|---|---|---|
| Provider | DeepSeek | Tencent |
| Noometry Index | 39.1 | 43.6 |
| Released | 2025-12-01 | 2026-07-06 |
| Weights | Open | Open |
| Context window | 128K | 262K |
| Max output | 128K | 128K |
| Input $ / M tokens | $0.58 | $0.13 |
| Output $ / M tokens | $1.68 | $0.53 |
| Results tracked | 3 | 17 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Hy3 leads
DeepSeek-V3.2-Speciale: 39.9 (#136), Hy3: 46.6 (#60)
| Benchmark | DeepSeek-V3.2-Speciale | Hy3 |
|---|---|---|
| LMArena WebDev | — | 1509 |
| WeirdML | 46.7% | — |
| LMArena Coding | — | 1465 |
Reasoning DeepSeek-V3.2-Speciale leads
DeepSeek-V3.2-Speciale: 32.1 (#65), Hy3: 25.0 (#134)
| Benchmark | DeepSeek-V3.2-Speciale | Hy3 |
|---|---|---|
| SimpleBench | 52.6% | — |
| NYT Connections (extended) | — | 41.2% |
| LMArena Hard Prompts | — | 1449 |
Math Not comparable
DeepSeek-V3.2-Speciale: —, Hy3: 37.4 (#98)
| Benchmark | DeepSeek-V3.2-Speciale | Hy3 |
|---|---|---|
| LMArena Math | — | 1475 |
Knowledge Not comparable
DeepSeek-V3.2-Speciale: —, Hy3: 40.2 (#108)
| Benchmark | DeepSeek-V3.2-Speciale | Hy3 |
|---|---|---|
| LMArena Expert | — | 1461 |
Multilingual Not comparable
DeepSeek-V3.2-Speciale: —, Hy3: 53.7 (#60)
| Benchmark | DeepSeek-V3.2-Speciale | Hy3 |
|---|---|---|
| LMArena Non-English | — | 1427 |
| LMArena Chinese | — | 1496 |
| LMArena French | — | 1461 |
| LMArena Korean | — | 1390 |
| LMArena Russian | — | 1435 |
| LMArena Spanish | — | 1448 |
Instruction Following Not comparable
DeepSeek-V3.2-Speciale: —, Hy3: 75.3 (#68)
| Benchmark | DeepSeek-V3.2-Speciale | Hy3 |
|---|---|---|
| LMArena Instruction Following | — | 1425 |
Long Context Not comparable
DeepSeek-V3.2-Speciale: —, Hy3: 44.1 (#70)
| Benchmark | DeepSeek-V3.2-Speciale | Hy3 |
|---|---|---|
| LMArena Longer Query | — | 1443 |
Writing & Preference Hy3 leads
DeepSeek-V3.2-Speciale: 46.1 (#217), Hy3: 62.5 (#75)
| Benchmark | DeepSeek-V3.2-Speciale | Hy3 |
|---|---|---|
| LMArena Text | — | 1442 |
| LMArena Creative Writing | — | 1399 |
| EQ-Bench Creative Writing | 1276 | — |
| LMArena Multi-Turn | — | 1436 |
Frequently asked questions
Is DeepSeek-V3.2-Speciale better than Hy3?
Hy3 is the stronger model overall, scoring 43.6 to 39.1 on the Noometry Index.
Which is cheaper, DeepSeek-V3.2-Speciale or Hy3?
Hy3 is cheaper. It lists at $0.13 per million input tokens and $0.53 per million output tokens; DeepSeek-V3.2-Speciale lists at $0.58 and $1.68.
Is DeepSeek-V3.2-Speciale or Hy3 better for coding?
Hy3 scores higher on coding benchmarks: 46.6 versus 39.9 in the Noometry coding category.
Which has the bigger context window?
Hy3 does, with 262K tokens against 128K.
How many benchmarks do DeepSeek-V3.2-Speciale and Hy3 share?
0 benchmarks have published results for both models. DeepSeek-V3.2-Speciale has 3 scored results on Noometry and Hy3 has 17.