Model comparison
DeepSeek-V3.1-Terminus vs Hy3
Hy3 is the stronger model overall, scoring 44.2 to 43.1 on the Noometry Index.
Last verified . 10 shared benchmarks.
DeepSeek
TencentSummary
- They share 10 benchmarks with published results for both. DeepSeek-V3.1-Terminus scores higher in 1 category and Hy3 in 6 categories; 5 gaps are clear of the uncertainty.
- The widest gap is in coding, where Hy3 leads 46.8 to 42.0.
- Hy3 is cheaper at $0.0825 / $0.33 per million input/output tokens, against $0.27 / $1 for DeepSeek-V3.1-Terminus.
- Hy3 accepts more context: 262K tokens versus 164K.
Side by side
| DeepSeek-V3.1-Terminus | Hy3 | |
|---|---|---|
| Provider | DeepSeek | Tencent |
| Noometry Index | 43.1 | 44.2 |
| Released | 2025-09-22 | 2026-07-06 |
| Weights | Open | Open |
| Context window | 164K | 262K |
| Max output | 147K | 128K |
| Input $ / M tokens | $0.27 | $0.0825 |
| Output $ / M tokens | $1 | $0.33 |
| Results tracked | 16 | 19 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Hy3 leads
DeepSeek-V3.1-Terminus: 42.0 (#113), Hy3: 46.8 (#63)
| Benchmark | DeepSeek-V3.1-Terminus | Hy3 |
|---|---|---|
| LMArena Coding | 1426 | 1464 |
| LMArena WebDev | — | 1508 |
| SciCode | 40.6% | — |
| ALE-Bench | 745.17 | — |
Reasoning Too close to call
DeepSeek-V3.1-Terminus: 26.4 (#133), Hy3: 26.1 (#136)
| Benchmark | DeepSeek-V3.1-Terminus | Hy3 |
|---|---|---|
| LMArena Hard Prompts | 1426 | 1447 |
| Kagi LLM Benchmark | 57.4% | — |
| NYT Connections (extended) | — | 41.2% |
| CritPt | 1.7% | — |
| DTBench | 81.3% | — |
| LMCA | 28.6% | — |
Math Hy3 leads
DeepSeek-V3.1-Terminus: 38.5 (#137), Hy3: 40.1 (#93)
| Benchmark | DeepSeek-V3.1-Terminus | Hy3 |
|---|---|---|
| LMArena Math | 1402 | 1475 |
Knowledge Not comparable
DeepSeek-V3.1-Terminus: —, Hy3: 40.8 (#114)
| Benchmark | DeepSeek-V3.1-Terminus | Hy3 |
|---|---|---|
| LMArena Expert | — | 1460 |
Multilingual Hy3 leads
DeepSeek-V3.1-Terminus: 52.1 (#92), Hy3: 53.5 (#65)
| Benchmark | DeepSeek-V3.1-Terminus | Hy3 |
|---|---|---|
| LMArena Non-English | 1407 | 1426 |
| LMArena Russian | 1436 | 1432 |
| LMArena Chinese | — | 1493 |
| LMArena French | — | 1461 |
| LMArena German | — | 1439 |
| LMArena Japanese | — | 1392 |
| LMArena Korean | — | 1395 |
| LMArena Spanish | — | 1456 |
Instruction Following Hy3 leads
DeepSeek-V3.1-Terminus: 74.0 (#106), Hy3: 75.1 (#70)
| Benchmark | DeepSeek-V3.1-Terminus | Hy3 |
|---|---|---|
| LMArena Instruction Following | 1404 | 1426 |
Long Context Too close to call
DeepSeek-V3.1-Terminus: 43.4 (#97), Hy3: 44.1 (#75)
| Benchmark | DeepSeek-V3.1-Terminus | Hy3 |
|---|---|---|
| LMArena Longer Query | 1421 | 1442 |
Writing & Preference Hy3 leads
DeepSeek-V3.1-Terminus: 61.0 (#92), Hy3: 62.2 (#81)
| Benchmark | DeepSeek-V3.1-Terminus | Hy3 |
|---|---|---|
| LMArena Text | 1419 | 1439 |
| LMArena Creative Writing | 1403 | 1402 |
| LMArena Multi-Turn | 1411 | 1436 |
Frequently asked questions
Is DeepSeek-V3.1-Terminus better than Hy3?
Hy3 is the stronger model overall, scoring 44.2 to 43.1 on the Noometry Index.
Which is cheaper, DeepSeek-V3.1-Terminus or Hy3?
Hy3 is cheaper. It lists at $0.0825 per million input tokens and $0.33 per million output tokens; DeepSeek-V3.1-Terminus lists at $0.27 and $1.
Is DeepSeek-V3.1-Terminus or Hy3 better for coding?
Hy3 scores higher on coding benchmarks: 46.8 versus 42.0 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-Terminus and Hy3 share?
10 benchmarks have published results for both models. DeepSeek-V3.1-Terminus has 16 scored results on Noometry and Hy3 has 19.