MMaDA 8B MixCoT needs about 9.8 GB of memory at Q4_K_M (a 5.0 GB file) with an 8K context window, so it fits on a 12 GB GPU. This repository has 25 quants, from 0.0 GB to 6.6 GB (Q6_K).
MMaDA 8B MixCoT – a multimodal diffusion foundation model
MMaDA 8B MixCoT is a 8.08 billion‑parameter llada architecture trained on 4096‑token context. It supports multimodal reasoning, chain‑of‑thought generation, and text‑to‑image tasks. The model is released under the MIT license.
Good at
- Multimodal reasoning
- Chain‑of‑thought generation
- Text‑to‑image synthesis
Limits
- No vision support in GGUF
- Requires 8 k context for full performance
- No explicit language list
Best for
- Research on multimodal diffusion models
- Developing multimodal assistants
- Testing chain‑of‑thought generation
Who it’s for. Researchers and developers interested in multimodal diffusion foundations. The model is suitable for experimentation and prototyping in environments that support GGUF and llama.cpp.
Quants & memory
| Quant | File | Est. memory | vs VRAM | Result | |
|---|---|---|---|---|---|
| — | 0.0 GB | 2.7 GB | Fits | Get file → | |
| IQ1_S | 2.1 GB | 4.7 GB | Fits | Get file → | |
| IQ1_M | 2.2 GB | 4.8 GB | Fits | Get file → | |
| IQ2_XXS | 2.4 GB | 5.1 GB | Fits | Get file → | |
| IQ2_XS | 2.6 GB | 5.3 GB | Fits | Get file → | |
| IQ2_S | 2.8 GB | 5.5 GB | Fits | Get file → | |
| Q2_K_S | 3.0 GB | 5.6 GB | Fits | Get file → | |
| IQ2_M | 3.0 GB | 5.6 GB | Fits | Get file → | |
| Q2_K | 3.2 GB | 5.9 GB | Fits | Get file → | |
| IQ3_XXS | 3.3 GB | 5.9 GB | Fits | Get file → | |
| IQ3_XS | 3.5 GB | 6.2 GB | Fits | Get file → | |
| IQ3_S | 3.7 GB | 6.3 GB | Fits | Get file → | |
| Q3_K_S | 3.7 GB | 6.3 GB | Fits | Get file → | |
| IQ3_M | 3.9 GB | 6.5 GB | Fits | Get file → | |
| Q3_K_M | 4.1 GB | 6.7 GB | Fits | Get file → | |
| Q3_K_L | 4.4 GB | 7.0 GB | Fits | Get file → | |
| IQ4_XS | 4.5 GB | 7.1 GB | Fits | Get file → | |
| IQ4_NL | 4.7 GB | 7.3 GB | Tight | Get file → | |
| Q4_0 | 4.7 GB | 7.4 GB | Tight | Get file → | |
| Q4_K_S | 4.7 GB | 7.4 GB | Tight | Get file → | |
| Q4_K_M | 5.0 GB | 7.6 GB | Tight | Get file → | |
| Q4_1 | 5.2 GB | 7.8 GB | Tight | Get file → | |
| Q5_K_S | 5.6 GB | 8.3 GB | Too big | Get file → | |
| Q5_K_M | 5.8 GB | 8.4 GB | Too big | Get file → | |
| Q6_K | 6.6 GB | 9.3 GB | Too big | Get file → |
For 8 GB VRAM, use IQ1_S (file size 2.06 GB, memory 6.9 GB). For 12 GB VRAM, Q4_K_M (file size 4.97 GB, memory 9.8 GB) is a good balance. For 16 GB VRAM, Q5_K_M (file size 5.78 GB, memory 10.6 GB) offers higher precision.
What GPU do you need for MMaDA 8B MixCoT?
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: the best quant that fits is the main file (0.0 GB file, about 4.8 GB in use).
- 8 GB VRAM: the best quant that fits is IQ3_XXS (3.3 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).
Which MMaDA 8B MixCoT quant should you choose?
Q4_K_M is the usual starting point for MMaDA 8B MixCoT: it balances file size and quality well. Here it is a 5.0 GB file.
If memory is tight, the smallest file is the first one in the table at 0.0 GB, with the quality trade-offs that come with fewer bits.
New to quant names? Read GGUF quantization explained.
Specifications
Architecture from the GGUF header
Source
Frequently asked questions about MMaDA 8B MixCoT
How much VRAM does MMaDA 8B MixCoT need?
About 9.8 GB at Q4_K_M with an 8K context window. That is the 5.0 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 MMaDA 8B MixCoT on an 8 GB GPU?
Yes, with IQ3_XXS: the 3.3 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 MMaDA 8B MixCoT GGUF?
It depends on the quant: file is 0.0 GB, IQ1_S is 2.1 GB, IQ1_M is 2.2 GB, IQ2_XXS is 2.4 GB, IQ2_XS is 2.6 GB, IQ2_S is 2.8 GB and 19 more.
What context length does MMaDA 8B MixCoT 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 MMaDA 8B MixCoT GGUF?
The files are on Hugging Face in the mradermacher/MMaDA-8B-MixCoT-i1-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 MMaDA 8B MixCoT 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 MMaDA 8B MixCoT on 8 GB VRAM?
Yes, the IQ1_S quant uses 6.9 GB at 8 k context, fitting within 8 GB VRAM.
Does the model support vision tasks?
The GGUF version does not include vision support; it focuses on text and diffusion generation.
What license does the model use?
The model is released under the MIT license.