SUPERBRAIN-1-G4
The public card describes this as the best run of SUPERBRAIN-1’s G-line (g4_morph) on google/gemma-4-E4B (Apache-2.0): a smaller, multimodal-capable sibling of the Fanar-based SUPERBRAIN-1. It is a completion model, not a chat model. Published files include a Q5_K_M GGUF, a LoRA adapter, and an image projector from a sibling run that the card says has not been graded on images. It was not in the 18 Oct 2026 catalog snapshot; it appeared on the public Hugging Face org list when we re-checked on 22 Nov 2026.
- Type
- model
- Task / pipeline
- text-generation
- License
- apache-2.0
- Size
- Gemma-4-E4B base (GGUF)
- Local or cloud
- Local open-weights (download from Hugging Face)
- Catalog snapshot (18 Oct 2026)
- not in that snapshot
- How we found it
- Live public list: GET https://huggingface.co/api/models?author=abuz8studios (not in the 18 Oct 2026 catalog snapshot).
- HF API (2026-11-22T12:30:06-05:00)
- downloads 30; likes 0; last modified 2026-10-14T13:23:44.000Z; repo size 6.6 GiB
- Base model (from the card)
- google/gemma-4-E4B
Sources: ABUZ8 catalog snapshot data/catalog.json (generated 2026-10-18T05:46Z), and where noted the public Hugging Face API plus the model card, fetched 2026-11-22T12:30:06-05:00. Unknown fields are labeled unknown. “Run in HQ” opens HQ so you can connect a local or API model; it does not auto-load these weights.