Skip to content
Other

One-For-All-4B-GGUF

quantized model files by mradermacher

2 min read

This repository provides static quantizations (GGUF builds) for the multimodal model TendieLabs/One-For-All-4B. The GGUF format lets you run language and multimodal models on consumer hardware (such as a standard PC or laptop using llama.cpp and compatible tools). Available compression options range from fast, compact files to higher-precision builds (including Q4_K_M, Q8_0, IQ4_XS, and others), alongside files for the multimodal projector (mmproj). It is a great fit for developers and hobbyists looking to run this architecture locally with minimal VRAM requirements.

Key features

  • Full suite of static GGUF quantizations (including Q2_K, Q3_K_M, Q4_K_M, Q8_0, and more)
  • Multimodal projector companion files included (mmproj-f16 and mmproj-Q8_0)
  • Optimized for local inference on standard consumer hardware
  • Multiple compression levels to balance output quality and file size

What you can do

  • Run the model locally using compatible clients and libraries (such as llama.cpp)
  • Pick the right quantization size to fit your available RAM or VRAM
  • Integrate multimodal capabilities into local data processing pipelines

Example

Using these GGUF files follows the standard llama.cpp workflow. Point your preferred client or CLI utility to the downloaded model file and its matching mmproj file to get started.

Read moreDetails
Base model: TendieLabs/One-For-All-4B
Library: transformers / GGUF
Available formats: x-f16, Q4_K_S, Q2_K, Q8_0, Q6_K, Q3_K_M, Q3_K_S, Q3_K_L, Q4_K_M, Q5_K_S, Q5_K_M, IQ4_XS
Multimodal components: mmproj-Q8_0 (0.5 GB), mmproj-f16 (0.8 GB)
Discuss Telegram 0

Will it run?

Runs in GGUF F16

The most precise of the ready quantised files fits.

  • GGUF F16weights 7.8 GB · ready files9.1 GB
  • GGUF Q8_0recommendedweights 4.2 GB · ready files5.1 GB
  • GGUF Q6_Kweights 3.2 GB · ready files4.0 GB
  • GGUF Q5_K_Mweights 2.9 GB · ready files3.7 GB
  • GGUF Q5_K_Sweights 2.8 GB · ready files3.6 GB
  • GGUF Q4_K_Mweights 2.5 GB · ready files3.3 GB
  • GGUF Q4_K_Sweights 2.4 GB · ready files3.1 GB
  • GGUF IQ4_XSweights 2.4 GB · ready files3.1 GB
  • GGUF Q3_K_Lweights 2.3 GB · ready files3.0 GB
  • GGUF Q3_K_Mweights 2.1 GB · ready files2.8 GB
  • GGUF Q3_K_Sweights 1.9 GB · ready files2.6 GB
  • GGUF Q2_Kweights 1.8 GB · ready files2.5 GB

Tip: go 8-bit.GGUF Q8_0 is within about a percent of full precision in quality, takes half the memory and runs faster on RTX 40/50 and H100. Full precision only pays off for fine-tuning.

What to get for this model

Massed ComputeOpen Massed Compute

Your own GPU: GGUF Q8_0 — from RTX 2060 · GGUF F16 — from RTX 3080

Massed Compute hourly prices, November 2026. The discounted price applies with the coupon at checkout.

Estimate: weights plus working memory. Long videos, high resolutions and long contexts need more.

Run it on Massed Compute

The script downloads the model to the rented machine. How to run it is in the model card.

  1. 1. Rent a machine with a fitting GPU

    Open Massed Compute
  2. 2. Paste the script into the machine's terminal

    Connect over SSH (the Massed Compute panel shows the command) or open a terminal on the machine's desktop.

    HF_TOKEN / CIVITAI_TOKENThe script starts with token lines: paste yours between the quotes. The Hugging Face token is for gated models, the Civitai one for Civitai models.

    # mradermacher/One-For-All-4B-GGUF
    mkdir -p ~/models
    curl -LsSf https://astral.sh/uv/install.sh | sh
    source "$HOME/.local/bin/env"
    uv tool install "huggingface_hub[cli]"
    hf download mradermacher/One-For-All-4B-GGUF --local-dir ~/models/One-For-All-4B-GGUF
    # The model is in ~/models. How to run it is in the model card (link above).
    Add more LoRAs to One-For-All-4B-GGUF
  3. 3. Use it

    The model is in ~/models. How to run it is in the model card (link above).

Stop the machine in the panel when you are done: it bills by the minute while it runs.

How to speed it up

Picked for this model type. Choose your hardware above to see what fits you exactly.

  • FlashAttention 44.0.0b40from November 18, 2026new release

    fastest exact attention on B200

    Kernels for Blackwell (beta). Exact attention without quantisation — fine for everything, LLMs included.

    How to enable: pip install flash-attn-4; vLLM and SGLang pick it up themselves.

    Earlier generations (2)
    • FlashAttention 3 — The Hopper (H100/H200) version with FP8.
    • FlashAttention 2 — The standard for RTX 30/40/50 and A100: exact attention without the memory cost of long contexts.
  • torch.compile2.15.0from October 28, 2026new release

    10–30% for free

    Compiles the model for your GPU. Slower first run, faster afterwards.

    How to enable: model = torch.compile(model); in ComfyUI the TorchCompileModel node.

Model family

  1. Qwen3.5-4B-Basefamily root
  2. Qwen3.5-4B
  3. One-For-All-4B
  4. One-For-All-4B-GGUF

New posts about versions of Qwen3.5-4B-Base (quants, finetunes, LoRAs) will show up under the bell.

Quantised versions 12All on Hugging Face
Fine-tunes 12All on Hugging Face
LoRAs and adapters 12All on Hugging Face
Merges 1All on Hugging Face

About the model

mradermacher/One-For-All-4B-GGUF
Author
mradermacher
Category
Other
Released
Posted

On NeyroLearn

Hugging Face

00

Ask AI about this model

Answered by Gemini 4 · the chat is kept only in this tab

Comments