Skip to content
zephel01 / AbyssView on GitHub
Run 50-280 GB MoE LLMs (DeepSeek V4/V4.1 Flash, GLM-5.3-Flash, Qwen3.8-Flash-Next, Ornith) on one consumer GPU: experts spread over VRAM, RAM and NVMe SSD. Browser chat + OpenAI/Anthropic APIs. CUDA, ROCm, DGX Spark.
☆40Nov 16, 2026Updated this week

Alternatives and similar repositories for Abyss

Users that are interested in Abyss are comparing it to the libraries listed below. We may earn a commission when you buy through links labeled 'Ad' on this page.

Sorting:

Are these results useful?