schneewolflabs/B1-9B-GGUF overview
B1 9B — GGUF Q8 0 the quant every card number was measured on and vision mmproj for schneewolflabs/B1 9B https://huggingface.co/schneewolflabs/B1 9B — the B0 9…
Runs locally from ~879.0 MB disk (4 GB VRAM class GPUs with llama.cpp / guIDE).
Repository Files & Downloads
Model Details
| Model ID | schneewolflabs/B1-9B-GGUF |
|---|---|
| Author | schneewolflabs |
| Pipeline | — |
| License | apache-2.0 |
| Base model | schneewolflabs/B1-9B |
| Last modified | 2026-09-06T04:47:02.000Z |
Model README
---
base_model: [schneewolflabs/B1-9B]
license: apache-2.0
language: [en]
tags: [gguf, agents, tool-use, reasoning, qwen3.5]
---
B1-9B — GGUF
Q8_0 (the quant every card number was measured on) and vision mmproj for
schneewolflabs/B1-9B — the B0-9B iteration
that answers after it thinks.
llama-server -m B1-9B-Q8_0.gguf -ngl 99 -c 8192 --jinja -fa on -np 1 \
--spec-type draft-mtp --spec-draft-n-max 4 \
--mmproj B1-9B-mmproj-f16.ggufSame architecture and layer count as B0-9B, so the
B0-9B-GGUF deployment guide — offload rules,
f16 KV cache, 6GB recipes — applies unchanged. With thinking on, pass tool definitions through
the native tools field and do not prefix /think; that request shape is where the
answer-after-thinking gain is largest.
Run schneewolflabs/B1-9B-GGUF with guIDE
Download guIDE — the AI-native code editor with local LLM inference and 69 built-in tools.
Source: Hugging Face · Compare models