Skip to content
Other

clairic-flash-GGUF

static quants of the Cloudflare model

2 min read

The repository offers a set of quantized versions of the Cloudflare/clairic-flash model in GGUF format (format for efficiently running large language models on consumer hardware). Quantization (compressing model weights to save memory) is done in various options ranging from compact Q2_K to high-quality Q8_0 and f16. The set is intended for developers and enthusiasts who want to run the original model locally with lower VRAM and RAM requirements.

Key features

  • Wide selection of compression options from 4.1 GB to 18.5 GB.
  • Recommended fast configurations Q4_K_S and Q4_K_M.
  • Availability of ultra-compact formats like Q2_K and IQ4_XS for constrained systems.
  • The GGUF format is compatible with popular local runtimes (such as llama.cpp).

What you can do

  • Run the model locally in a terminal or via compatible UI interfaces.
  • Test different quantization levels to find the balance between generation speed and quality.
  • Integrate the model into local text processing pipelines on modest hardware.

Example

To run it, download the desired quantization file (for example, Q4_K_M) and use it in any GGUF-compatible client or library that supports the format by specifying the path to the weights file.

Read moreDetails
Base model: Cloudflare/clairic-flash
Available file sizes: from 4.1 GB (Q2_K) to 18.5 GB (f16)
Recommended options: Q4_K_M (5.9 GB) and Q4_K_S (5.6 GB)
Format: GGUF static quants
Discuss Telegram 0

Will it run?

Runs in GGUF F16

The most precise of the ready quantised files fits.

  • GGUF F16weights 17 GB · ready files19 GB
  • GGUF Q8_0recommendedweights 9.1 GB · ready files11 GB
  • GGUF Q6_Kweights 7.0 GB · ready files8.2 GB
  • GGUF Q5_K_Mweights 6.2 GB · ready files7.3 GB
  • GGUF Q5_K_Sweights 6.0 GB · ready files7.1 GB
  • GGUF Q4_K_Mweights 5.4 GB · ready files6.5 GB
  • GGUF Q4_K_Sweights 5.1 GB · ready files6.1 GB
  • GGUF IQ4_XSweights 5.0 GB · ready files6.1 GB
  • GGUF Q3_K_Lweights 4.8 GB · ready files5.8 GB
  • GGUF Q3_K_Mweights 4.5 GB · ready files5.4 GB
  • GGUF Q3_K_Sweights 4.1 GB · ready files5.0 GB
  • GGUF Q2_Kweights 3.7 GB · ready files4.6 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 2080 Ti · GGUF F16 — from RTX 4090

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/clairic-flash-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/clairic-flash-GGUF --local-dir ~/models/clairic-flash-GGUF
    # The model is in ~/models. How to run it is in the model card (link above).
    Add more LoRAs to clairic-flash-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. Qwen3.5-9B-Basefamily root
  2. Qwen3.5-9B
  3. clairic-flash
  4. clairic-flash-GGUF

New posts about versions of Qwen3.5-9B-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 6All on Hugging Face
Merges 1All on Hugging Face

About the model

mradermacher/clairic-flash-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