Model comparison
Gemini 3.1 Pro Preview vs Hy3
Gemini 3.1 Pro Preview is the stronger model overall, scoring 56.7 to 44.2 on the Noometry Index. Hy3 costs 31× less per token, which makes it the better buy when Gemini 3.1 Pro Preview's lead doesn't matter for your workload.
Last verified . 19 shared benchmarks.
Google
TencentSummary
- They share 19 benchmarks with published results for both. Gemini 3.1 Pro Preview scores higher in 7 categories and Hy3 in 1 category; 8 gaps are clear of the uncertainty.
- The widest gap is in reasoning, where Gemini 3.1 Pro Preview leads 71.7 to 26.1.
- The biggest single-benchmark swing is NYT Connections (extended): 97.4% for Gemini 3.1 Pro Preview and 41.2% for Hy3.
- Hy3 is cheaper at $0.0825 / $0.33 per million input/output tokens, against $2 / $12 for Gemini 3.1 Pro Preview.
- Gemini 3.1 Pro Preview accepts more context: 1.05M tokens versus 262K.
- Hy3 has downloadable open weights; the other is API-only.
Side by side
| Gemini 3.1 Pro Preview | Hy3 | |
|---|---|---|
| Provider | Tencent | |
| Noometry Index | 56.7 | 44.2 |
| Released | 2026-02-19 | 2026-07-06 |
| Weights | Proprietary | Open |
| Context window | 1.05M | 262K |
| Max output | 66K | 128K |
| Input $ / M tokens | $2 | $0.0825 |
| Output $ / M tokens | $12 | $0.33 |
| Results tracked | 71 | 19 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Hy3 leads
Gemini 3.1 Pro Preview: 42.5 (#99), Hy3: 46.8 (#63)
| Benchmark | Gemini 3.1 Pro Preview | Hy3 |
|---|---|---|
| LMArena WebDev | 1447 | 1508 |
| LMArena Coding | 1484 | 1464 |
| SWE-bench Verified | 75.6% | — |
| DeepSWE | 11.7% | — |
| SciCode | 58.9% | — |
| GSO | 22.6% | — |
| WeirdML | 72.1% | — |
| MirrorCode | 8.9% | — |
| ALE-Bench | 1,161 | — |
| AlgoTune | 2.02 | — |
Agentic & Tool Use Not comparable
Gemini 3.1 Pro Preview: 37.7 (#34), Hy3: —
| Benchmark | Gemini 3.1 Pro Preview | Hy3 |
|---|---|---|
| Terminal-Bench | 80.2% | — |
| APEX-Agents | 35.3% | — |
| τ²-bench Banking | 26% | — |
| DeepResearch Bench | 47.8% | — |
| PostTrainBench | 22% | — |
| BALROG | 57% | — |
| ExploitBench | 26.1% | — |
| GBAEval | 0.8% | — |
| GDP.pdf | 17% | — |
| LMArena Search | 1211 | — |
| METR Time Horizons | 77% | — |
| Vending-Bench 2 | 3,774 | — |
Reasoning Gemini 3.1 Pro Preview leads
Gemini 3.1 Pro Preview: 71.7 (#12), Hy3: 26.1 (#136)
| Benchmark | Gemini 3.1 Pro Preview | Hy3 |
|---|---|---|
| NYT Connections (extended) | 97.4% | 41.2% |
| LMArena Hard Prompts | 1485 | 1447 |
| ARC-AGI-2 | 77.1% | — |
| SimpleBench | 79.6% | — |
| ARC-AGI-1 | 98% | — |
| CritPt | 17.7% | — |
| Chess Puzzles | 55% | — |
| EnigmaEval | 36.8% | — |
| Thematic Generalization | 79.4% | — |
| EBR-Bench | 14.3% | — |
| Mystery Game Puzzles | 34% | — |
| DTBench | 97.1% | — |
| LMCA | 53.8% | — |
| Epoch Capabilities Index | 154.77 | — |
| ForecastBench | 59 | — |
Math Gemini 3.1 Pro Preview leads
Gemini 3.1 Pro Preview: 62.1 (#34), Hy3: 40.1 (#93)
| Benchmark | Gemini 3.1 Pro Preview | Hy3 |
|---|---|---|
| LMArena Math | 1485 | 1475 |
| FrontierMath (Tiers 1-3) | 59.6% | — |
| FrontierMath Tier 4 | 26.8% | — |
| MathArena Final-Answer Competitions | 86.5% | — |
| OTIS Mock AIME 2024-2025 | 95.6% | — |
| ProofBench | 26% | — |
| FrontierMath (Feb 2025 set) | 36.9% | — |
| FrontierMath Tier 4 (v1) | 16.7% | — |
Knowledge Gemini 3.1 Pro Preview leads
Gemini 3.1 Pro Preview: 71.8 (#3), Hy3: 40.8 (#114)
| Benchmark | Gemini 3.1 Pro Preview | Hy3 |
|---|---|---|
| LMArena Expert | 1485 | 1460 |
| GPQA Diamond | 94.4% | — |
| Humanity's Last Exam | 46.4% | — |
| SimpleQA Verified | 73.5% | — |
| Vectara Hallucination Rate | 10.4% | — |
Multimodal Not comparable
Gemini 3.1 Pro Preview: 37.9 (#69), Hy3: —
| Benchmark | Gemini 3.1 Pro Preview | Hy3 |
|---|---|---|
| LMArena Vision | 1296 | — |
| Blueprint-Bench 2 | 26.5% | — |
| Furniture Assembly | 26.7% | — |
| LMArena Document | 1444 | — |
Multilingual Gemini 3.1 Pro Preview leads
Gemini 3.1 Pro Preview: 57.0 (#12), Hy3: 53.5 (#65)
| Benchmark | Gemini 3.1 Pro Preview | Hy3 |
|---|---|---|
| LMArena Non-English | 1477 | 1426 |
| LMArena Chinese | 1529 | 1493 |
| LMArena French | 1487 | 1461 |
| LMArena German | 1491 | 1439 |
| LMArena Japanese | 1493 | 1392 |
| LMArena Korean | 1455 | 1395 |
| LMArena Russian | 1498 | 1432 |
| LMArena Spanish | 1479 | 1456 |
Instruction Following Gemini 3.1 Pro Preview leads
Gemini 3.1 Pro Preview: 77.0 (#32), Hy3: 75.1 (#70)
| Benchmark | Gemini 3.1 Pro Preview | Hy3 |
|---|---|---|
| LMArena Instruction Following | 1466 | 1426 |
Long Context Gemini 3.1 Pro Preview leads
Gemini 3.1 Pro Preview: 47.4 (#18), Hy3: 44.1 (#75)
| Benchmark | Gemini 3.1 Pro Preview | Hy3 |
|---|---|---|
| LMArena Longer Query | 1483 | 1442 |
| CL-bench | 20.8% | — |
| CL-bench Life | 16.9% | — |
Writing & Preference Gemini 3.1 Pro Preview leads
Gemini 3.1 Pro Preview: 66.1 (#37), Hy3: 62.2 (#81)
| Benchmark | Gemini 3.1 Pro Preview | Hy3 |
|---|---|---|
| LMArena Text | 1481 | 1439 |
| LMArena Creative Writing | 1482 | 1402 |
| LMArena Multi-Turn | 1488 | 1436 |
| EQ-Bench Creative Writing | 1491 | — |
| EQ-Bench 4 | 1142 | — |
Frequently asked questions
Is Gemini 3.1 Pro Preview better than Hy3?
Gemini 3.1 Pro Preview is the stronger model overall, scoring 56.7 to 44.2 on the Noometry Index. Hy3 costs 31× less per token, which makes it the better buy when Gemini 3.1 Pro Preview's lead doesn't matter for your workload.
Which is cheaper, Gemini 3.1 Pro Preview or Hy3?
Hy3 is cheaper. It lists at $0.0825 per million input tokens and $0.33 per million output tokens; Gemini 3.1 Pro Preview lists at $2 and $12.
Is Gemini 3.1 Pro Preview or Hy3 better for coding?
Hy3 scores higher on coding benchmarks: 46.8 versus 42.5 in the Noometry coding category.
Which has the bigger context window?
Gemini 3.1 Pro Preview does, with 1.05M tokens against 262K.
How many benchmarks do Gemini 3.1 Pro Preview and Hy3 share?
19 benchmarks have published results for both models. Gemini 3.1 Pro Preview has 71 scored results on Noometry and Hy3 has 19.