Running 276B Inkling Small on a Single Mac Studio via nvfp4 Quantization
pcuenq · x · 2026-07-31
A developer demonstrated how to efficiently deploy the newly released Inkling Small (276B parameters) within the Apple ecosystem.
- Massively Reduced Requirements: Using the official nvfp4 checkpoint, the model requires only 150 GB of memory to run on a single Mac Studio, or it can be split across two 128GB laptops.
- More Aggressive Quantization: The author noted that with more aggressive quantization, the model should fit entirely on a single laptop soon.
- Tooling: A PR has been submitted on GitHub to natively support nvfp4 and mxfp4 models, enabling direct inference via mlx-vlm.
Related event: Inkling-Small: 276B params, 12B active, open-source tops ARC(48 posts)→
More from Infra
- NVIDIA launches RTX Spark for local AI on Windows PCs — ostrisai · 2026-10-16
- SkyRL ships Day-0 RL support for Beam with +62% LoRA decode throughput in vLLM — vllm_project · 2026-10-16
- DFlash2 delivers up to 3.03x decode-only speedup for Beam across five benchmarks — vllm_project · 2026-10-16
- vLLM's Beam serving kernels cut GB200 forward time from 8.15 to 6.47 ms at TP8 batch 128 — vllm_project · 2026-10-16
- Reflection's 501B MoE Beam hits 463 tok/s on 8x GB200 with NVFP4 at day 0 — zhyncs42 · 2026-10-16
- Gemma 4 31B QAT on one 24GB card: 46 tok/s plain, ~80 with draft model on RTX 4090 — Subject_Poet_3861 · 2026-10-16