Kev 27b needs about 18.5 GB of memory at Q4_K_M (a 15.8 GB file) with an 8K context window, so it fits on a 24 GB GPU. This repository has 11 quants, from 10.3 GB (Q2_K) to 27.2 GB (Q8_0). Longer context needs more: at 32K tokens the KV cache alone adds about 8.6 GB.
Kev 27b – a decision model for typed document scoring
Kev 27b is a decision model that reads a state and a set of typed questions, returning calibrated probabilities for each option. It is designed for classification, routing, triage, and policy checks on documents up to 65,536 tokens. The model is a full‑weight fine‑tune of Qwen3.8‑27B.
Good at
- Typed decision scoring
- Confidence‑based workflow automation
- Integration with System One API
Limits
- No text generation or summarisation
- Requires at least 80 GB GPU or 96 GB RAM
- Only English states up to 65,536 tokens
Best for
- Document triage systems
- Automated routing with human review
- Policy compliance checks
Who it’s for. Developers building systems that need reliable, thresholdable probabilities for document decisions. The model works best in environments with large GPU memory or high‑end Apple Silicon setups.
Quants & memory
| Quant | File | Est. memory | vs VRAM | Result | |
|---|---|---|---|---|---|
| Q2_K | 10.3 GB | 12.9 GB | Too big | Get file → | |
| Q3_K_S | 11.5 GB | 14.2 GB | Too big | Get file → | |
| Q3_K_M | 12.8 GB | 15.4 GB | Too big | Get file → | |
| Q3_K_L | 13.8 GB | 16.4 GB | Too big | Get file → | |
| IQ4_XS | 14.5 GB | 17.2 GB | Too big | Get file → | |
| Q4_K_S | 14.9 GB | 17.5 GB | Too big | Get file → | |
| Q4_K_M | 15.8 GB | 18.5 GB | Too big | Get file → | |
| Q5_K_S | 17.8 GB | 20.5 GB | Too big | Get file → | |
| Q5_K_M | 18.4 GB | 21.0 GB | Too big | Get file → | |
| Q6_K | 21.0 GB | 23.7 GB | Too big | Get file → | |
| Q8_0 | 27.2 GB | 29.9 GB | Too big | Get file → |
For 8 GB VRAM use Q2_K (10.29 GB). For 16 GB VRAM use Q3_K_S (11.53 GB). For 24 GB VRAM use Q4_K_M (15.83 GB). Memory estimates are approximate and depend on context length.
What GPU do you need for Kev 27b?
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: the best quant that fits is Q3_K_M (12.8 GB file, about 15.4 GB in use).
- 24 GB VRAM: the best quant that fits is Q6_K (21.0 GB file, about 23.7 GB in use).
- 32 GB VRAM: the best quant that fits is Q8_0 (27.2 GB file, about 29.9 GB in use).
Which Kev 27b quant should you choose?
Q4_K_M is the usual starting point for Kev 27b: it balances file size and quality well. Here it is a 15.8 GB file.
The Q8_0 file is 27.2 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.
If memory is tight, the smallest file is Q2_K at 10.3 GB, with the quality trade-offs that come with fewer bits. Use it only when memory is very tight. Expect a clear drop in quality, which is usually worse on small models and easier to live with on large ones.
New to quant names? Read GGUF quantization explained.
Specifications
Architecture from the GGUF header
Source
Frequently asked questions about Kev 27b
How much VRAM does Kev 27b need?
About 18.5 GB at Q4_K_M with an 8K context window. That is the 15.8 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 Kev 27b on an 8 GB GPU?
Not at an 8K context. The smallest file needs about 12.9 GB, so you would need a larger GPU or CPU offloading.
What is the file size of Kev 27b GGUF?
It depends on the quant: Q2_K is 10.3 GB, Q3_K_S is 11.5 GB, Q3_K_M is 12.8 GB, Q3_K_L is 13.8 GB, IQ4_XS is 14.5 GB, Q4_K_S is 14.9 GB and 5 more.
What context length does Kev 27b 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 Kev 27b GGUF?
The files are on Hugging Face in the mradermacher/kev-27b-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 Kev 27b 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 Kev 27b on 8 GB VRAM?
Yes, the Q2_K quant is 10.29 GB and requires about 12.9 GB of VRAM at 8k context.
Does Kev 27b support image inputs?
No, the model only processes text or JSON states and does not support vision.
What license applies to Kev 27b?
The model is released under the apache-2.0 license.