JiRackUltra_32b needs about 22.5 GB of memory at Q4_K_M (a 19.9 GB file) with an 8K context window, so it fits on a 24 GB GPU. This repository has 7 quants, from 12.3 GB (Q2_K) to 65.5 GB. Longer context needs more: at 32K tokens the KV cache alone adds about 8.6 GB.
JiRackUltra_32b – a 32B CPU‑optimized Qwen2 model with ternary support
JiRackUltra_32b is a 32‑billion‑parameter Qwen2 model optimized for CPU inference. It uses BitNet ternary quantization and a custom tokenizer that adds Routing, Media, Vision, Sound, Tool call, and Robotics tags. The model is released under the MIT license.
Good at
- High‑quality CPU inference with QAT
- Advanced tool calling and agentic use
- Large context support up to 131072 tokens
Limits
- No vision support in current quantizations
- Requires 8‑k context for memory estimates
- Large model size limits deployment on low‑VRAM GPUs
Best for
- RAG deployments on cloud CPUs
- Agent coding IDEs like Cursor or Windsurf
- Tool‑calling applications in robotics or routing
Who it’s for. Developers building CPU‑centric inference pipelines who need a large, efficient model with advanced tool‑calling capabilities. The model is suitable for cloud‑ready deployments and agentic applications.
Quants & memory
| Quant | File | Est. memory | vs VRAM | Result | |
|---|---|---|---|---|---|
| Q2_K | 12.3 GB | 15.0 GB | Too big | Get file → | |
| Q3_K_M | 15.9 GB | 18.6 GB | Too big | Get file → | |
| Q4_K_M | 19.9 GB | 22.5 GB | Too big | Get file → | |
| Q5_K_M | 23.3 GB | 25.9 GB | Too big | Get file → | |
| Q6_K | 26.9 GB | 29.5 GB | Too big | Get file → | |
| Q8_0 | 34.8 GB | 37.5 GB | Too big | Get file → | |
| — | 65.5 GB | 68.2 GB | Too big | Get file → |
For 8‑GB VRAM, Q2_K (12.31 GB) or Q3_K_M (15.94 GB) are recommended. If more memory is available, Q4_K_M (19.85 GB) or Q5_K_M (23.26 GB) provide higher precision. The Q6_K and Q8_0 variants require larger VRAM and are best for high‑performance setups.
What GPU do you need for JiRackUltra_32b?
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 Q2_K (12.3 GB file, about 15.0 GB in use).
- 24 GB VRAM: the best quant that fits is Q4_K_M (19.9 GB file, about 22.5 GB in use).
- 32 GB VRAM: the best quant that fits is Q6_K (26.9 GB file, about 29.5 GB in use).
- 48 GB VRAM: the best quant that fits is Q8_0 (34.8 GB file, about 37.5 GB in use).
Which JiRackUltra_32b quant should you choose?
Q4_K_M is the usual starting point for JiRackUltra_32b: it balances file size and quality well. Here it is a 19.9 GB file.
The Q8_0 file is 34.8 GB, about 1.8 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 12.3 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 JiRackUltra_32b
How much VRAM does JiRackUltra_32b need?
About 22.5 GB at Q4_K_M with an 8K context window. That is the 19.9 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 JiRackUltra_32b on an 8 GB GPU?
Not at an 8K context. The smallest file needs about 15.0 GB, so you would need a larger GPU or CPU offloading.
What is the file size of JiRackUltra_32b GGUF?
It depends on the quant: Q2_K is 12.3 GB, Q3_K_M is 15.9 GB, Q4_K_M is 19.9 GB, Q5_K_M is 23.3 GB, Q6_K is 26.9 GB, Q8_0 is 34.8 GB and 1 more.
What context length does JiRackUltra_32b support?
The GGUF header lists a trained context length of 131,072 tokens (128K). 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 JiRackUltra_32b GGUF?
The files are on Hugging Face in the CMSManhattan/JiRackUltra_32b repository. The download buttons on this page open the file on Hugging Face. ggufmodels.com does not host any model files.
What license does JiRackUltra_32b use?
Hugging Face lists the license as mit. 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 JiRackUltra_32b on 8 GB VRAM?
Yes, the Q2_K or Q3_K_M quantizations fit within 8 GB VRAM with approximate memory usage of 15–18.6 GB at 8k context.
Does the model support vision tasks?
No, the current quantizations do not include vision support.
What license applies to JiRackUltra_32b?
The model is released under the MIT license.