Model comparison
DeepSeek-V3 vs Hy3
Hy3 is the stronger model overall, scoring 44.2 to 39.5 on the Noometry Index.
Last verified . 17 shared benchmarks.
DeepSeek
TencentSummary
- They share 17 benchmarks with published results for both. DeepSeek-V3 scores higher in 0 categories and Hy3 in 8 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in long context, where Hy3 leads 44.1 to 34.0.
- Hy3 is cheaper at $0.0825 / $0.33 per million input/output tokens, against $0.24 / $0.90 for DeepSeek-V3.
- Hy3 accepts more context: 262K tokens versus 164K.
Side by side
| DeepSeek-V3 | Hy3 | |
|---|---|---|
| Provider | DeepSeek | Tencent |
| Noometry Index | 39.5 | 44.2 |
| Released | 2024-12-26 | 2026-07-06 |
| Weights | Open | Open |
| Context window | 164K | 262K |
| Max output | 164K | 128K |
| Input $ / M tokens | $0.24 | $0.0825 |
| Output $ / M tokens | $0.90 | $0.33 |
| Results tracked | 60 | 19 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Hy3 leads
DeepSeek-V3: 42.3 (#106), Hy3: 46.8 (#63)
| Benchmark | DeepSeek-V3 | Hy3 |
|---|---|---|
| LMArena Coding | 1368 | 1464 |
| Aider Polyglot | 55.1% | — |
| LMArena WebDev | — | 1508 |
| SciCode | 35.8% | — |
| WeirdML | 36.1% | — |
| BigCodeBench Instruct | 50% | — |
| LiveBench Coding | 70.9% | — |
| BigCodeBench Complete | 62.2% | — |
| HumanEval+ | 86.6% | — |
| MBPP+ | 73% | — |
Agentic & Tool Use Not comparable
DeepSeek-V3: —, Hy3: —
| Benchmark | DeepSeek-V3 | Hy3 |
|---|---|---|
| METR Time Horizons | 49.6% | — |
Reasoning Hy3 leads
DeepSeek-V3: 20.5 (#236), Hy3: 26.1 (#136)
| Benchmark | DeepSeek-V3 | Hy3 |
|---|---|---|
| LMArena Hard Prompts | 1365 | 1447 |
| SimpleBench | 27.2% | — |
| Kagi LLM Benchmark | 52.3% | — |
| NYT Connections (extended) | — | 41.2% |
| CritPt | 0% | — |
| LiveBench Reasoning | 65.8% | — |
| DTBench | 64.8% | — |
| LiveBench Data Analysis | 60.9% | — |
| LMCA | 15.5% | — |
| BIG-Bench Hard | 87.5% | — |
| Epoch Capabilities Index | 135.94 | — |
| ForecastBench | 59.1 | — |
| HellaSwag | 88.9% | — |
| LiveBench | 66.9% | — |
| PIQA | 84.7% | — |
| WinoGrande | 85.2% | — |
Math Hy3 leads
DeepSeek-V3: 32.1 (#219), Hy3: 40.1 (#93)
| Benchmark | DeepSeek-V3 | Hy3 |
|---|---|---|
| LMArena Math | 1373 | 1475 |
| OTIS Mock AIME 2024-2025 | 37.8% | — |
| Omni-MATH | 40.3% | — |
| LiveBench Math | 73.5% | — |
| MATH Level 5 | 75.5% | — |
| FrontierMath (Feb 2025 set) | 1.7% | — |
Knowledge Hy3 leads
DeepSeek-V3: 37.5 (#155), Hy3: 40.8 (#114)
| Benchmark | DeepSeek-V3 | Hy3 |
|---|---|---|
| LMArena Expert | 1351 | 1460 |
| GPQA Diamond | 67.6% | — |
| MMLU-Pro | 72.3% | — |
| Confabulations | 26.1% | — |
| Vectara Hallucination Rate | 6.1% | — |
| GPQA (HELM) | 53.8% | — |
| ARC (AI2) Challenge | 95.3% | — |
| MMLU | 87.2% | — |
| TriviaQA | 82.9% | — |
Multilingual Hy3 leads
DeepSeek-V3: 48.5 (#143), Hy3: 53.5 (#65)
| Benchmark | DeepSeek-V3 | Hy3 |
|---|---|---|
| LMArena Non-English | 1358 | 1426 |
| LMArena Chinese | 1391 | 1493 |
| LMArena French | 1385 | 1461 |
| LMArena German | 1374 | 1439 |
| LMArena Japanese | 1333 | 1392 |
| LMArena Korean | 1319 | 1395 |
| LMArena Russian | 1373 | 1432 |
| LMArena Spanish | 1358 | 1456 |
Instruction Following Hy3 leads
DeepSeek-V3: 72.8 (#130), Hy3: 75.1 (#70)
| Benchmark | DeepSeek-V3 | Hy3 |
|---|---|---|
| LMArena Instruction Following | 1345 | 1426 |
| LiveBench Instruction Following | 81.5% | — |
| IFEval | 83.2% | — |
Long Context Hy3 leads
DeepSeek-V3: 34.0 (#253), Hy3: 44.1 (#75)
| Benchmark | DeepSeek-V3 | Hy3 |
|---|---|---|
| LMArena Longer Query | 1352 | 1442 |
| Fiction.LiveBench | 50% | — |
Writing & Preference Hy3 leads
DeepSeek-V3: 57.4 (#130), Hy3: 62.2 (#81)
| Benchmark | DeepSeek-V3 | Hy3 |
|---|---|---|
| LMArena Text | 1375 | 1439 |
| LMArena Creative Writing | 1364 | 1402 |
| LMArena Multi-Turn | 1389 | 1436 |
| Short-Story Creative Writing | 77% | — |
| EQ-Bench Creative Writing | 1472 | — |
| WildBench | 83% | — |
| LiveBench Language | 49.1% | — |
Frequently asked questions
Is DeepSeek-V3 better than Hy3?
Hy3 is the stronger model overall, scoring 44.2 to 39.5 on the Noometry Index.
Which is cheaper, DeepSeek-V3 or Hy3?
Hy3 is cheaper. It lists at $0.0825 per million input tokens and $0.33 per million output tokens; DeepSeek-V3 lists at $0.24 and $0.90.
Is DeepSeek-V3 or Hy3 better for coding?
Hy3 scores higher on coding benchmarks: 46.8 versus 42.3 in the Noometry coding category.
Which has the bigger context window?
Hy3 does, with 262K tokens against 164K.
How many benchmarks do DeepSeek-V3 and Hy3 share?
17 benchmarks have published results for both models. DeepSeek-V3 has 60 scored results on Noometry and Hy3 has 19.