LLM EXPLORER 59,516 MODELS INDEXED

Hy3 MLX Q9 by inferencerlabs

By inferencerlabs · 1016 downloads
HF Repo ↗ Copy URL

Hy3 MLX Q9 is an open-source language model by inferencerlabs. Features: LLM, VRAM: 330.9GB, Context: 256K, Quantized, LLM Explorer Score: 0.29.

Base model:quantized:tencent/h...   Base model:tencent/hy3   Conversational   En   Hy v3   Mlx   Q9   Quantized   Region:us   Sharded   Tensorflow
Model Card on HF 🤗: https://huggingface.co/inferencerlabs/Hy3-MLX-Q9 

Hy3 MLX Q9 Parameters and Internals

LLM NameHy3 MLX Q9
Repository 🤗https://huggingface.co/inferencerlabs/Hy3-MLX-Q9 
Base Model(s)  Hy3   tencent/Hy3
Required VRAM330.9 GB
Updated2026-08-08
Maintainerinferencerlabs
Model Typehy_v3
Model Files  10.6 GB: 1-of-34   9.7 GB: 2-of-34   9.8 GB: 3-of-34   9.7 GB: 4-of-34   9.7 GB: 5-of-34   9.8 GB: 6-of-34   9.7 GB: 7-of-34   9.7 GB: 8-of-34   9.8 GB: 9-of-34   9.7 GB: 10-of-34   9.7 GB: 11-of-34   9.8 GB: 12-of-34   9.7 GB: 13-of-34   9.7 GB: 14-of-34   9.8 GB: 15-of-34   9.7 GB: 16-of-34   9.7 GB: 17-of-34   9.8 GB: 18-of-34   9.7 GB: 19-of-34   9.7 GB: 20-of-34   9.8 GB: 21-of-34   9.7 GB: 22-of-34   9.7 GB: 23-of-34   9.8 GB: 24-of-34   9.7 GB: 25-of-34   9.7 GB: 26-of-34   9.8 GB: 27-of-34   9.7 GB: 28-of-34   9.7 GB: 29-of-34   9.8 GB: 30-of-34   9.7 GB: 31-of-34   9.7 GB: 32-of-34   9.8 GB: 33-of-34   8.8 GB: 34-of-34
Supported Languagesen
Quantization Typeq9
Model ArchitectureHYV3ForCausalLM
Context Length262144
Model Max Length262144
Transformers Version5.6.0
Tokenizer ClassTokenizersBackend
Padding Token<｜hy_pad:opensource｜>
Vocabulary Size120832