Edge0/Edge0-35B-A3B-preview — new model trending #30 on Hugging Face
Edge0 released a 35B sparse MoE model running in under 3 GiB of memory at 15 tok/s via SSD expert offload and int4 quantization.
Edge0-35b-a3b-preview is a 35B-parameter MoE (256 experts, 4 active per token) built on Qwen3.5-MoE 35B-A3B, shipped as a 4-bit checkpoint with LoRA and prerouter adapters under Apache 2.0. The edge0 framework streams expert weights from SSD on demand, bounding peak active memory at 2.9 GiB and achieving 14.9-17.7 tok/s decode on a Mac mini M4 Pro (MLX backend). Recover-LoRA distillation keeps the int4 model within 3.9 points of its fp16 base (79.2 vs 83.2 average on OpenCompass benchmarks including AIME 2026, HumanEval, GPQA-Diamond, MMLU-Pro, and IFBench).
- 35B MoE (256 experts, 4 active) based on Qwen3.5-MoE 35B-A3B, Apache 2.0
- SSD expert offload bounds peak active RAM at 2.9 GiB with 4-bit weights
- Prerouter predicts expert routing ahead, adding up to +59% decode throughput
- 15 tok/s decode and 140 tok/s warm prefill on Mac mini M4 Pro (MLX)
- int4 + Recover-LoRA scores 79.2 avg vs 83.2 for the fp16 base
- Vendors
- Edge0
- Products
- Edge0-35B-A3B-previewedge0
- Organizations
- Edge0-AI
From the original article
3 min read at huggingface.cotext-generation · 951 downloads · 123 likes · license apache-2. 0
By Edge0 · huggingface.co
A short excerpt; the article and its images belong to the publisher. Summary, key facts and assessment above are AI-written from the article and can be wrong.