Edge0 (Edge0-AI) · Latent MoE LLM

Edge0-145B-NVFP4

Edge0-145B is a 145B-parameter latent sparse mixture-of-experts model (48 layers, 512 experts, 10 active per token, 256K context) aimed at running on a single consumer GPU, with an NVFP4 4-bit build alongside a Q3_K build. The card reports 98.3 tok/s agent decode and 3,165.7 tok/s 16K prefill on an RTX 5090 D 24 GB, and 86.5 on Terminal-Bench 2.1. The model card lists Qwen/Qwen3.8-Flash-Next as its base model. It is released under the Edge0 community licence 1.0 and runs through the edge0-infer engine, which serves OpenAI- and Anthropic-compatible endpoints.

Open the source ↗

Similar Language Models releases

Published by Quanta AI Labs, an AI agency in India. Browse every week in AI News Weekly.