StartLux-Decision-v1.1-27B-Instant needs about 19.2 GB of memory at Q4_K_M (a 16.5 GB file) with an 8K context window, so it fits on a 24 GB GPU. This repository has 2 quants, from 16.5 GB (Q4_K_M) to 28.6 GB (Q8_0). Longer context needs more: at 32K tokens the KV cache alone adds about 8.6 GB.
27B decision model that returns probabilities for choices, yes/no, or scores without text generation
StartLux-Decision-v1.1-27B-Instant is a 27‑billion‑parameter decision model. It accepts a state (text or JSON with up to eight images) and typed questions, then outputs probabilities for each option in a single forward pass. The model is optimized for decision tasks and uses bidirectional prompt attention.
Good at
- Fast single‑pass probability outputs
- Handles up to 8 images per state
- Compatible with System One API
Limits
- Requires 19–31 GB VRAM for 8 k context
- Vision encoder separate (928 MB)
- CC‑BY‑NC‑4.0 license
Best for
- Automated decision support
- Rapid multi‑choice evaluation
- Embedded systems with limited compute
Who it’s for. Developers building decision‑making tools that need quick probability estimates without text generation. The model is suitable for applications requiring structured outputs and compatibility with the System One API.
Quants & memory
| Quant | File | Est. memory | vs VRAM | Result | |
|---|---|---|---|---|---|
| Q4_K_M | 16.5 GB | 19.2 GB | Too big | Get file → | |
| Q8_0 | 28.6 GB | 31.2 GB | Too big | Get file → |
For 8 k context, Q4_K_M needs about 19.2 GB VRAM and is 16.55 GB on disk. Q8_0 uses 31.2 GB VRAM and is 28.6 GB on disk. Choose Q4_K_M if memory is tight; use Q8_0 for higher accuracy.
What GPU do you need for StartLux-Decision-v1.1-27B-Instant?
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: no quant fits at an 8K context.
- 8 GB VRAM: no quant fits at an 8K context.
- 12 GB VRAM: no quant fits at an 8K context.
- 16 GB VRAM: no quant fits at an 8K context.
- 24 GB VRAM: the best quant that fits is Q4_K_M (16.5 GB file, about 19.2 GB in use).
- 32 GB VRAM: the best quant that fits is Q8_0 (28.6 GB file, about 31.2 GB in use).
Which StartLux-Decision-v1.1-27B-Instant quant should you choose?
Q4_K_M is the usual starting point for StartLux-Decision-v1.1-27B-Instant: it balances file size and quality well. Here it is a 16.5 GB file.
The Q8_0 file is 28.6 GB, about 1.7 times the size of Q4_K_M. It keeps the weights closer to the original, which only matters if you have the memory to spare.
New to quant names? Read GGUF quantization explained.
Specifications
Architecture from the GGUF header
Source
Frequently asked questions about StartLux-Decision-v1.1-27B-Instant
How much VRAM does StartLux-Decision-v1.1-27B-Instant need?
About 19.2 GB at Q4_K_M with an 8K context window. That is the 16.5 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 StartLux-Decision-v1.1-27B-Instant on an 8 GB GPU?
Not at an 8K context. The smallest file needs about 19.2 GB, so you would need a larger GPU or CPU offloading.
What is the file size of StartLux-Decision-v1.1-27B-Instant GGUF?
It depends on the quant: Q4_K_M is 16.5 GB, Q8_0 is 28.6 GB.
What context length does StartLux-Decision-v1.1-27B-Instant support?
The GGUF header lists a trained context length of 262,144 tokens (256K). The KV cache grows with the context you use: about 2.1 GB at 8K and 8.6 GB at 32K.
Where can I download StartLux-Decision-v1.1-27B-Instant GGUF?
The files are on Hugging Face in the startlux-models/StartLux-Decision-v1.1-27B-Instant-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 StartLux-Decision-v1.1-27B-Instant use?
Hugging Face lists the license as cc-by-nc-4.0. 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 this model on 8 GB VRAM?
No. The smallest quantized file requires at least 19 GB VRAM for 8 k context.
Does the model support image input?
Yes, up to eight images per state via a separate 928 MB vision encoder.
What license applies?
The model is released under CC‑BY‑NC‑4.0.