Model comparison
DeepSeek-V3.1 vs Hy3
Hy3 is the stronger model overall, scoring 44.2 to 42.8 on the Noometry Index.
Last verified . 17 shared benchmarks.
DeepSeek
TencentSummary
- They share 17 benchmarks with published results for both. DeepSeek-V3.1 scores higher in 2 categories and Hy3 in 6 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in long context, where Hy3 leads 44.1 to 36.3.
- Hy3 is cheaper at $0.13 / $0.53 per million input/output tokens, against $0.25 / $0.95 for DeepSeek-V3.1.
- Hy3 accepts more context: 262K tokens versus 164K.
Side by side
| DeepSeek-V3.1 | Hy3 | |
|---|---|---|
| Provider | DeepSeek | Tencent |
| Noometry Index | 42.8 | 44.2 |
| Released | 2025-08-21 | 2026-07-06 |
| Weights | Open | Open |
| Context window | 164K | 262K |
| Max output | 8K | 128K |
| Input $ / M tokens | $0.25 | $0.13 |
| Output $ / M tokens | $0.95 | $0.53 |
| Results tracked | 27 | 19 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Hy3 leads
DeepSeek-V3.1: 40.3 (#144), Hy3: 46.8 (#63)
| Benchmark | DeepSeek-V3.1 | Hy3 |
|---|---|---|
| LMArena Coding | 1417 | 1464 |
| LMArena WebDev | — | 1508 |
| WeirdML | 38.4% | — |
Reasoning DeepSeek-V3.1 leads
DeepSeek-V3.1: 27.9 (#110), Hy3: 26.1 (#136)
| Benchmark | DeepSeek-V3.1 | Hy3 |
|---|---|---|
| LMArena Hard Prompts | 1417 | 1447 |
| SimpleBench | 40% | — |
| Kagi LLM Benchmark | 53.2% | — |
| NYT Connections (extended) | — | 41.2% |
| DTBench | 82.7% | — |
| LMCA | 24.3% | — |
| Epoch Capabilities Index | 139.92 | — |
| ForecastBench | 58 | — |
Math Hy3 leads
DeepSeek-V3.1: 38.9 (#122), Hy3: 40.1 (#93)
| Benchmark | DeepSeek-V3.1 | Hy3 |
|---|---|---|
| LMArena Math | 1420 | 1475 |
Knowledge DeepSeek-V3.1 leads
DeepSeek-V3.1: 43.7 (#90), Hy3: 40.8 (#114)
| Benchmark | DeepSeek-V3.1 | Hy3 |
|---|---|---|
| LMArena Expert | 1405 | 1460 |
| Vectara Hallucination Rate | 5.5% | — |
Multilingual Hy3 leads
DeepSeek-V3.1: 51.6 (#106), Hy3: 53.5 (#65)
| Benchmark | DeepSeek-V3.1 | Hy3 |
|---|---|---|
| LMArena Non-English | 1400 | 1426 |
| LMArena Chinese | 1469 | 1493 |
| LMArena French | 1447 | 1461 |
| LMArena German | 1411 | 1439 |
| LMArena Japanese | 1378 | 1392 |
| LMArena Korean | 1337 | 1395 |
| LMArena Russian | 1405 | 1432 |
| LMArena Spanish | 1431 | 1456 |
Instruction Following Hy3 leads
DeepSeek-V3.1: 73.9 (#110), Hy3: 75.1 (#70)
| Benchmark | DeepSeek-V3.1 | Hy3 |
|---|---|---|
| LMArena Instruction Following | 1400 | 1426 |
Long Context Hy3 leads
DeepSeek-V3.1: 36.3 (#232), Hy3: 44.1 (#75)
| Benchmark | DeepSeek-V3.1 | Hy3 |
|---|---|---|
| LMArena Longer Query | 1422 | 1442 |
| Fiction.LiveBench | 52.8% | — |
Writing & Preference Hy3 leads
DeepSeek-V3.1: 60.3 (#98), Hy3: 62.2 (#81)
| Benchmark | DeepSeek-V3.1 | Hy3 |
|---|---|---|
| LMArena Text | 1420 | 1439 |
| LMArena Creative Writing | 1401 | 1402 |
| LMArena Multi-Turn | 1408 | 1436 |
| EQ-Bench Creative Writing | 1436 | — |
Frequently asked questions
Is DeepSeek-V3.1 better than Hy3?
Hy3 is the stronger model overall, scoring 44.2 to 42.8 on the Noometry Index.
Which is cheaper, DeepSeek-V3.1 or Hy3?
Hy3 is cheaper. It lists at $0.13 per million input tokens and $0.53 per million output tokens; DeepSeek-V3.1 lists at $0.25 and $0.95.
Is DeepSeek-V3.1 or Hy3 better for coding?
Hy3 scores higher on coding benchmarks: 46.8 versus 40.3 in the Noometry coding category.
Which has the bigger context window?
Hy3 does, with 262K tokens against 164K.
How many benchmarks do DeepSeek-V3.1 and Hy3 share?
17 benchmarks have published results for both models. DeepSeek-V3.1 has 27 scored results on Noometry and Hy3 has 19.