---
title: Edge0 35B Preview — Showcase
emoji: 🧠
colorFrom: blue
colorTo: indigo
sdk: static
sdk_version: 4.0.0
app_file: index.html
pinned: false
license: apache-2.0
short_description: 35B MoE LLM that runs in ~3 GiB — animated demo
---

# Edge0-35B-A3B-preview — Showcase

A static showcase page for the [Edge0-35B-A3B-preview](https://huggingface.co/Edge0/Edge0-35B-A3B-preview) model:
35B-class sparse MoE that streams experts from SSD in under 3 GiB of active memory.

Includes an interactive simulation of the streaming-inference pipeline (expert routing map, prerouter queue,
active-memory meter, token decode), animated benchmarks vs the fp16 teacher, architecture details and
run-it-on-your-Mac instructions.

Link: https://huggingface.co/spaces/rcode21212/edge0-35b-showcase