K2 Horizon MoVA 36B A4B DERISKED needs about 24.5 GB of memory at Q4_K_M (a 22.4 GB file) with an 8K context window, so it fits on a 32 GB GPU. This repository has 16 quants, from 0.1 GB (Q8_0) to 39.9 GB (Q8_0). Longer context needs more: at 32K tokens the KV cache alone adds about 6.4 GB.
K2 Horizon MoVA 36B A4B DERISKED – 37.4B parameter LLM in GGUF format
K2 Horizon MoVA 36B A4B DERISKED is a 37.4B parameter model with 4B active parameters. It is a mixture of experts and supports a context of 524288 tokens. The model is released under the Apache‑2.0 license.
Good at
- Large context handling
- Mixture of experts architecture
- Generation with multiple quantizations
Limits
- Requires 8‑k context for memory estimates
- Not all quantizations produce readable text
- No vision support
Best for
- Long‑form content generation
- Research prototyping
- High‑capacity chat applications
Who it’s for. Developers and researchers needing a large‑parameter model with extensive context. The model is suitable for environments that can accommodate the memory requirements of the chosen quantization.
Quants & memory
| Quant | File | Est. memory | vs VRAM | Result | |
|---|---|---|---|---|---|
| Q8_0 | 0.1 GB | 2.2 GB | Fits | Get file → | |
| TQ1_0 | 8.6 GB | 10.7 GB | Too big | Get file → | |
| TQ2_0 | 10.3 GB | 12.4 GB | Too big | Get file → | |
| IQ3_M | 16.6 GB | 18.7 GB | Too big | Get file → | |
| IQ3_M | 16.7 GB | 18.8 GB | Too big | Get file → | |
| IQ4_XS | 20.3 GB | 22.4 GB | Too big | Get file → | |
| Q4_0 | 21.3 GB | 23.4 GB | Too big | Get file → | |
| IQ4_NL | 21.4 GB | 23.5 GB | Too big | Get file → | |
| Q4_K_S | 21.4 GB | 23.5 GB | Too big | Get file → | |
| MXFP4_MOE | 21.7 GB | 23.9 GB | Too big | Get file → | |
| Q4_K_M | 22.4 GB | 24.5 GB | Too big | Get file → | |
| Q4_K | 22.8 GB | 24.9 GB | Too big | Get file → | |
| Q4_1 | 23.6 GB | 25.7 GB | Too big | Get file → | |
| Q5_K_M | 26.5 GB | 28.6 GB | Too big | Get file → | |
| Q6_K | 30.8 GB | 32.9 GB | Too big | Get file → | |
| Q8_0 | 39.9 GB | 42.0 GB | Too big | Get file → |
For 8‑GB VRAM, use Q4_K_M (22.39 GB file, ~24.5 GB memory at 8 k context). For 16‑GB VRAM, Q5_K_M (26.46 GB file, ~28.6 GB memory) or Q6_K (30.79 GB file, ~32.9 GB memory). Larger VRAM allows Q8_0 (39.85 GB file, ~42 GB memory).
What GPU do you need for K2 Horizon MoVA 36B A4B DERISKED?
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: the best quant that fits is Q8_0 (0.1 GB file, about 2.2 GB in use).
- 6 GB VRAM: the best quant that fits is Q8_0 (0.1 GB file, about 2.2 GB in use).
- 8 GB VRAM: the best quant that fits is Q8_0 (0.1 GB file, about 2.2 GB in use).
- 12 GB VRAM: the best quant that fits is TQ1_0 (8.6 GB file, about 10.7 GB in use).
- 16 GB VRAM: the best quant that fits is TQ2_0 (10.3 GB file, about 12.4 GB in use).
- 24 GB VRAM: the best quant that fits is MXFP4_MOE (21.7 GB file, about 23.9 GB in use).
- 32 GB VRAM: the best quant that fits is Q5_K_M (26.5 GB file, about 28.6 GB in use).
- 48 GB VRAM: the best quant that fits is Q8_0 (39.9 GB file, about 42.0 GB in use).
Which K2 Horizon MoVA 36B A4B DERISKED quant should you choose?
Q4_K_M is the usual starting point for K2 Horizon MoVA 36B A4B DERISKED: it balances file size and quality well. Here it is a 22.4 GB file.
The Q8_0 file is 0.1 GB, about 0.0 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 Q8_0 at 0.1 GB, with the quality trade-offs that come with fewer bits. Close to lossless. It costs roughly twice the memory of a 4-bit file, so it suits cases where you have plenty of VRAM.
New to quant names? Read GGUF quantization explained.
Specifications
Architecture from the GGUF header
Source
Frequently asked questions about K2 Horizon MoVA 36B A4B DERISKED
How much VRAM does K2 Horizon MoVA 36B A4B DERISKED need?
About 24.5 GB at Q4_K_M with an 8K context window. That is the 22.4 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 K2 Horizon MoVA 36B A4B DERISKED on an 8 GB GPU?
Yes, with Q8_0: the 0.1 GB file needs about 2.2 GB at 8K context. Larger quants in this repository do not fit in 8 GB.
What is the file size of K2 Horizon MoVA 36B A4B DERISKED GGUF?
It depends on the quant: Q8_0 is 0.1 GB, TQ1_0 is 8.6 GB, TQ2_0 is 10.3 GB, IQ3_M is 16.6 GB, IQ3_M is 16.7 GB, IQ4_XS is 20.3 GB and 10 more.
What context length does K2 Horizon MoVA 36B A4B DERISKED support?
The GGUF header lists a trained context length of 524,288 tokens (512K). The KV cache grows with the context you use: about 1.6 GB at 8K and 6.4 GB at 32K.
Where can I download K2 Horizon MoVA 36B A4B DERISKED GGUF?
The files are on Hugging Face in the freakyskittle/K2-Horizon-MoVA-36B-A4B-DERISKED-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 K2 Horizon MoVA 36B A4B DERISKED 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 this model on 8 GB VRAM?
Use the Q4_K_M quantization; it requires about 24.5 GB memory at 8 k context, which exceeds 8 GB VRAM. A smaller quant such as Q4_K_S would still exceed 8 GB.
Does the model support vision?
No, the model has no vision capabilities.
Are all quantizations usable for text generation?
TQ1_0 and TQ2_0 do not produce readable text; other tested quants generate readable output.