First MLX build of the viral DeepSeek-V4-Pro-Qwen3.5-9B-UNCENSORED-heretic — DeepSeek-V4-Pro reasoning distilled into Qwen3.5-9B, then decensored with the Heretic tool's Arbitrary-Rank Ablation. Runs on any Apple Silicon Mac with 16GB+ unified memory.
⚠️ Uncensored model. Refusal behaviour was removed upstream by abliteration (refusals 100/100 → 3/100, KL 0.042 from base). You are responsible for how you use it and for complying with applicable law.
| Architecture | Qwen3_5ForConditionalGeneration (text) |
| Layers | 32 (24 linear + 8 full attention) |
| Quantization | 4-bit affine, group 64 (from BF16 source) |
| Size | 5.6 GB |
| Context | 262,144 tokens |
| Vocab | 248,320 |
Not a coding specialist — no coding data in the distillation set.
| Memory resident | ~4.8 GB wired |
| Time to first token | ~1.9 s |
| Decode | ~32 tok/s |
Any MLX runtime: NovaMLX, mlx-lm, LM Studio.
mlx_lm.generate --model digiflow/DeepSeek-V4-Pro-Qwen3.5-9B-UNCENSORED-MLX-4bit --prompt "..."
model-00001-of-00002.safetensors 5.3 GBmodel-00002-of-00002.safetensors 600 MB