Can Hy3 run on the RTX 4060 Ti 16GB?
Not fully in VRAM. Even at 2-bit with 8K context, Hy3 needs about 102 GB, more than the 15.2 GB usable on the RTX 4060 Ti 16GB. It can still run with offload (slower): Q3_K_M with part of the model in system RAM, about 4.0–7.6 tokens/sec (estimate). Details and command.
Runs with offload (slower)
Runs with offload 3-bit (Q3_K_M) at 8K context: all attention layers plus the experts of 4 of 80 MoE layers on the RTX 4060 Ti 16GB (about 12.5 GB), and the experts of the other 76 layers in system RAM (about 118 GB). Needs a PC with at least 128 GB of system RAM. Estimated speed: 4.0–7.6 tokens/sec (slow but usable), versus a bandwidth ceiling of ~30 tok/s if it fit fully in VRAM.
llama-server -hf bartowski/Hy3-GGUF:Q3_K_M -c 8192 -ngl 999 --n-cpu-moe 76
Estimate, not a benchmark. The command downloads the GGUF from Hugging Face. The speed range spans dual-channel DDR4-3200 to DDR5-6000 system RAM; your CPU, RAM speed and llama.cpp version move it. If llama.cpp runs out of GPU memory, raise --n-cpu-moe by 1-2; if VRAM is left over, lower it for more speed. Recent llama.cpp builds can also pick the split automatically. How we estimate offload.
Memory by quant and context
| Quant | 4K ctx | 8K ctx | 32K ctx | 128K ctx | Max tok/s (est., 8K) |
|---|---|---|---|---|---|
| 2-bit (Q2_K) • | 101 GB No | 102 GB No | 110 GB No | 140 GB No | — |
| 3-bit (Q3_K_M) • | 130 GB No | 131 GB No | 139 GB No | 169 GB No | — |
| 4-bit (Q4_K_M) • | 172 GB No | 173 GB No | 180 GB No | 210 GB No | — |
| 5-bit (Q5_K_M) • | 200 GB No | 201 GB No | 209 GB No | 239 GB No | — |
| 6-bit (Q6_K) • | 241 GB No | 243 GB No | 250 GB No | 280 GB No | — |
| 8-bit (Q8_0) • | 298 GB No | 299 GB No | 306 GB No | 336 GB No | — |
| 16-bit (F16/BF16) | 558 GB No | 560 GB No | 567 GB No | 597 GB No | — |
• real GGUF file size from bartowski/Hy3-GGUF. Speed is a bandwidth ceiling (288 GB/s), not a benchmark. Methodology.
Hardware that does fit it
DGX SparkRyzen AI Max+ 395 128GBH200Mac M2 Ultra 192GBMac M3 Ultra 256GBMac M3 Ultra 512GB
What fits instead
- Cheapest listed GPU that runs Hy3: DGX Spark (128 GB, launch MSRP $3,999) at Q2_K, fully in VRAM at 8K context.
- Largest popular model that fits the RTX 4060 Ti 16GB: gpt-oss 20B (21B, MoE) at Q8_0, about 12.3 GB at 8K context.
"Runs" and "fits" mean the whole model in GPU memory at 8K context (4-bit or better for the model pick). Prices are US launch MSRPs, not today's street prices. "Popular" means the 60 most-downloaded models we track.
More
Hy3 on every GPU and Mac · Every model that runs on the RTX 4060 Ti 16GB · Other models that fit 16 GB VRAM · Try other settings in the calculator
Guides: Best GPU for local LLMs by budget · Can I run gpt-oss, Llama 70B or Qwen3? · Mac vs PC for local LLMs
Data: Hugging Face Hub (tencent/Hy3), NVIDIA specs. Model data last checked 2026-11-18; refreshed hourly.