LLaDA 1.5 needs about 9.7 GB of memory at Q4_K_M (a 4.9 GB file) with an 8K context window, so it fits on a 12 GB GPU. This repository has 12 quants, from 3.2 GB (Q2_K) to 16.0 GB (F16).
LLaDA 1.5 – 8.02B parameter diffusion language model
LLaDA 1.5 is an 8.02B parameter diffusion language model trained with variance‑reduced preference optimization. It supports a 4096 token context window and is released under the MIT license.
Good at
- Math tasks
- Code generation
- Alignment tasks
Limits
- No vision support
- Requires 8k context for full performance
- No mixture of experts
Best for
- Academic research
- Educational coding assistants
- Math problem solving
Who it’s for. Researchers and developers needing a diffusion‑based LLM with strong math and code capabilities. The model is open‑source and can be run locally with GGUF quantizations.
Quants & memory
| Quant | File | Est. memory | vs VRAM | Result | |
|---|---|---|---|---|---|
| Q2_K | 3.2 GB | 5.8 GB | Fits | Get file → | |
| Q3_K_S | 3.7 GB | 6.3 GB | Fits | Get file → | |
| Q3_K_M | 4.0 GB | 6.7 GB | Fits | Get file → | |
| Q3_K_L | 4.3 GB | 7.0 GB | Fits | Get file → | |
| IQ4_XS | 4.4 GB | 7.1 GB | Fits | Get file → | |
| Q4_K_S | 4.7 GB | 7.3 GB | Tight | Get file → | |
| Q4_K_M | 4.9 GB | 7.6 GB | Tight | Get file → | |
| Q5_K_S | 5.6 GB | 8.2 GB | Too big | Get file → | |
| Q5_K_M | 5.7 GB | 8.4 GB | Too big | Get file → | |
| Q6_K | 6.6 GB | 9.2 GB | Too big | Get file → | |
| Q8_0 | 8.5 GB | 11.2 GB | Too big | Get file → | |
| F16 | 16.0 GB | 18.7 GB | Too big | Get file → |
For 8 GB VRAM, use Q4_K_M (file size 4.93 GB, estimated 9.7 GB memory). For 16 GB VRAM, Q5_K_M (5.73 GB, 10.5 GB memory) or Q6_K (6.58 GB, 11.4 GB memory). Higher‑precision F16 (16.04 GB, 20.8 GB memory) offers best quality but needs more memory.
What GPU do you need for LLaDA 1.5?
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: the best quant that fits is Q2_K (3.2 GB file, about 8.0 GB in use).
- 12 GB VRAM: the best quant that fits is Q6_K (6.6 GB file, about 11.4 GB in use).
- 16 GB VRAM: the best quant that fits is Q8_0 (8.5 GB file, about 13.3 GB in use).
- 24 GB VRAM: the best quant that fits is F16 (16.0 GB file, about 20.8 GB in use).
Which LLaDA 1.5 quant should you choose?
Q4_K_M is the usual starting point for LLaDA 1.5: it balances file size and quality well. Here it is a 4.9 GB file.
The Q8_0 file is 8.5 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 3.2 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 LLaDA 1.5
How much VRAM does LLaDA 1.5 need?
About 9.7 GB at Q4_K_M with an 8K context window. That is the 4.9 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 LLaDA 1.5 on an 8 GB GPU?
Yes, with Q2_K: the 3.2 GB file needs about 8.0 GB at 8K context. Larger quants in this repository do not fit in 8 GB.
What is the file size of LLaDA 1.5 GGUF?
It depends on the quant: Q2_K is 3.2 GB, Q3_K_S is 3.7 GB, Q3_K_M is 4.0 GB, Q3_K_L is 4.3 GB, IQ4_XS is 4.4 GB, Q4_K_S is 4.7 GB and 6 more.
What context length does LLaDA 1.5 support?
The GGUF header lists a trained context length of 4,096 tokens (4K). The KV cache grows with the context you use: about 4.3 GB at 8K and 2.1 GB at 4K.
Where can I download LLaDA 1.5 GGUF?
The files are on Hugging Face in the mradermacher/LLaDA-1.5-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 LLaDA 1.5 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 LLaDA 1.5 on 8 GB VRAM?
Use the Q4_K_M quant; it requires about 9.7 GB memory at 8k context, which is close to 8 GB but may need swap.
Does LLaDA 1.5 support image inputs?
No, the model has no vision capabilities.