Hy3 models
3 Hy3 models from Tencent on Hugging Face, from 298.8B to 298.8B parameters, published by Tencent in BF16, with a 256K-token context. The smallest official model, Hy3, needs about 668 GB of VRAM at its published precision; the cheapest live fit is AMD MI300X at $10.52/hr.
Part of the Hunyuan series · Previous generation: Hunyuan Instruct · Next generation: Hy4
Pick a size
One row per official Hy3 size: the VRAM it needs at each precision, the cheapest GPU that holds it today, and the KV cache for a 32K-token context.
| Model | Parameters | Native VRAM | FP8 VRAM | INT4 VRAM | Live GPU fit (native) | Est. $/hr | KV cache at 32K |
|---|---|---|---|---|---|---|---|
| Hy3-preview | 298.8B | 668 GB | 334 GB | 167 GB | AMD MI300X × 4 | $10.52/hr | 10.0 GB |
VRAM is weights times a flat 1.2 overhead; the KV cache is a separate per-model figure at 16-bit, shown where the architecture is published. See the methodology.
Official models (2)
| Model | Parameters | Published as | VRAM needed | Live GPU fit | Est. $/hr |
|---|---|---|---|---|---|
| Hy3 | 298.8B | BF16 | 668 GB | AMD MI300X × 4 | $10.52/hr |
| Hy3-preview | 298.8B | BF16 | 668 GB | AMD MI300X × 4 | $10.52/hr |
Quantized versions (1)
| Model | Parameters | Published as | VRAM needed | Live GPU fit | Est. $/hr |
|---|---|---|---|---|---|
| Hy3-FP8 | 298.8B | F8_E4M3 | 334 GB | AMD MI300X × 2 | $5.26/hr |
VRAM is for the precision the model is published in: the weights times a flat 1.2 overhead, with the KV cache not included. See the methodology. The fit shown is the lowest-priced single GPU type that holds the model at that precision, or the lowest-priced multi-GPU set (up to 8) when none does.
More from Tencent
- tencent (3 models)
- Hunyuan Instruct (2 models)
- Hy-MT2 (3 models)
- Hy4 (2 models)