Clef-Flash needs about 5.7 GB of memory at Q3_K_S (a 4.1 GB file) with an 8K context window, so it fits on a 6 GB GPU. This repository has 4 quants, from 3.6 GB (IQ3_XXS) to 4.9 GB (IQ4_XS). Longer context needs more: at 32K tokens the KV cache alone adds about 4.3 GB.
Clef-Flash: 9B decision model for typed questions on text or images
Clef-Flash is a 9B multimodal decision model that reads state and schema of typed questions and returns probabilities for each option. It supports text, JSON, images, or video as input and outputs a single forward pass of logits. The model is post‑trained from Qwen/Qwen3.5‑9B and has no free‑form text generation.
Good at
- Typed decision making
- Multimodal input handling
- Fast single‑pass inference
Limits
- No free‑form text generation
- No vision support beyond images/video
- Requires llama.cpp 9 Oct 2026 or newer
Best for
- Decision support systems
- Automated form filling
- Rapid policy evaluation
Who it’s for. Developers building decision engines that need quick, probabilistic answers to typed questions. Ideal for applications that require deterministic outputs without text generation.
Quants & memory
| Quant | File | Est. memory | vs VRAM | Result | |
|---|---|---|---|---|---|
| IQ3_XXS | 3.6 GB | 5.2 GB | Fits | Get file → | |
| Q3_K_S | 4.1 GB | 5.7 GB | Fits | Get file → | |
| Q3_K_L | 4.5 GB | 6.1 GB | Fits | Get file → | |
| IQ4_XS | 4.9 GB | 6.5 GB | Fits | Get file → |
For 8 GB VRAM, IQ3_XXS (3.65 GB) uses 5.2 GB memory at 8 k context. Q3_K_S (4.14 GB) needs 5.7 GB memory. IQ3_K_L (4.52 GB) uses 6.1 GB memory. IQ4_XS (4.94 GB) requires 6.5 GB memory. Choose the smallest file that fits your VRAM budget.
What GPU do you need for Clef-Flash?
These figures add the file size, an fp16 KV cache at an 8K context window (worked out from this model’s own header) and about 0.5 GB for the runtime. They are estimates, so leave a little headroom, and expect to use less context or a smaller quant if you are right at the limit.
- 4 GB VRAM: no quant fits at an 8K context.
- 6 GB VRAM: the best quant that fits is Q3_K_S (4.1 GB file, about 5.7 GB in use).
- 8 GB VRAM: the best quant that fits is IQ4_XS (4.9 GB file, about 6.5 GB in use).
Which Clef-Flash quant should you choose?
Q3_K_S (4.1 GB) is the middle option in this repository. A reasonable choice when a 4-bit file will not fit, mostly on larger models. Quality loss is noticeable but often acceptable.
If memory is tight, the smallest file is IQ3_XXS at 3.6 GB, with the quality trade-offs that come with fewer bits. A reasonable choice when a 4-bit file will not fit, mostly on larger models. Quality loss is noticeable but often acceptable.
New to quant names? Read GGUF quantization explained.
Specifications
Architecture from the GGUF header
Source
Frequently asked questions about Clef-Flash
How much VRAM does Clef-Flash need?
About 5.7 GB at Q3_K_S with an 8K context window. That is the 4.1 GB file, plus the KV cache and roughly 0.5 GB of runtime overhead. Smaller quants need less and longer context needs more.
Can I run Clef-Flash on an 8 GB GPU?
Yes, with IQ4_XS: the 4.9 GB file needs about 6.5 GB at 8K context. Larger quants in this repository do not fit in 8 GB.
What is the file size of Clef-Flash GGUF?
It depends on the quant: IQ3_XXS is 3.6 GB, Q3_K_S is 4.1 GB, Q3_K_L is 4.5 GB, IQ4_XS is 4.9 GB.
What context length does Clef-Flash support?
The GGUF header lists a trained context length of 262,144 tokens (256K). The KV cache grows with the context you use: about 1.1 GB at 8K and 4.3 GB at 32K.
Where can I download Clef-Flash GGUF?
The files are on Hugging Face in the turintech/Clef-Flash-Artemis-GGUF repository. The download buttons on this page open the file on Hugging Face. ggufmodels.com does not host any model files.
What license does Clef-Flash use?
Hugging Face lists the license as apache-2.0. Licenses can differ between the quantized files and the original model, so check the repository and the base model before commercial use.
Can I run Clef-Flash on 8 GB VRAM?
Yes, the IQ3_XXS quantized file is 3.65 GB and requires about 5.2 GB memory at 8 k context, which fits in 8 GB VRAM.
Does Clef-Flash support image input?
Yes, it can read images or video as part of the state, but it does not generate text.