Qwen3.5-122B-A10B needs about 77.8 GB of memory at Q4_K_M (a 76.5 GB file) with an 8K context window, which is more than a 48 GB GPU holds. This repository has 24 quants, from 0.4 GB to 244.3 GB (BF16). Longer context needs more: at 32K tokens the KV cache alone adds about 3.2 GB.
122B‑parameter Qwen3.5 model with vision support
Qwen3.5‑122B‑A10B is a large causal language model with vision encoder. It contains 122 billion parameters, of which 10 billion are active in the Mixture‑of‑Experts. The model supports 262 144‑token context and is licensed under Apache‑2.0.
Good at
- Vision‑language tasks
- Long‑context reasoning
- Multilingual generation
Limits
- Requires high VRAM for full precision
- No native audio support
- Large file sizes for high‑quality quants
Best for
- Vision‑enabled chatbots
- Multimodal content creation
- Large‑scale inference on powerful GPUs
Who it’s for. Developers needing a multimodal foundation model for vision and text. The model is suitable for research and production environments that can accommodate its memory demands.
Quants & memory
| Quant | File | Est. memory | vs VRAM | Result | |
|---|---|---|---|---|---|
| — | 0.4 GB | 1.7 GB | Fits | Get file → | |
| IQ1_M | 34.2 GB | 35.5 GB | Too big | Get file → | |
| IQ2_XXS | 36.6 GB | 37.9 GB | Too big | Get file → | |
| IQ2_M | 39.1 GB | 40.5 GB | Too big | Get file → | |
| Q2_K_XL | 41.8 GB | 43.2 GB | Too big | Get file → | |
| IQ3_XXS | 44.7 GB | 46.1 GB | Too big | Get file → | |
| IQ3_S | 46.6 GB | 47.9 GB | Too big | Get file → | |
| Q3_K_S | 52.5 GB (3 parts) | 53.8 GB | Too big | Get file → | |
| Q3_K_M | 56.4 GB (3 parts) | 57.7 GB | Too big | Get file → | |
| Q3_K_XL | 57.0 GB (3 parts) | 58.3 GB | Too big | Get file → | |
| IQ4_XS | 60.2 GB (3 parts) | 61.5 GB | Too big | Get file → | |
| IQ4_NL | 61.4 GB (3 parts) | 62.7 GB | Too big | Get file → | |
| Q4_K_S | 71.7 GB (3 parts) | 73.0 GB | Too big | Get file → | |
| MXFP4_MOE | 74.7 GB (3 parts) | 76.0 GB | Too big | Get file → | |
| Q4_K_M | 76.5 GB (3 parts) | 77.8 GB | Too big | Get file → | |
| Q4_K_XL | 77.0 GB (3 parts) | 78.3 GB | Too big | Get file → | |
| Q5_K_S | 86.4 GB (3 parts) | 87.7 GB | Too big | Get file → | |
| Q5_K_M | 91.5 GB (3 parts) | 92.8 GB | Too big | Get file → | |
| Q5_K_XL | 91.9 GB (3 parts) | 93.2 GB | Too big | Get file → | |
| Q6_K | 101.0 GB (4 parts) | 102.3 GB | Too big | Get file → | |
| Q6_K_XL | 112.4 GB (4 parts) | 113.7 GB | Too big | Get file → | |
| Q8_0 | 129.9 GB (4 parts) | 131.2 GB | Too big | Get file → | |
| Q8_K_XL | 170.8 GB (5 parts) | 172.1 GB | Too big | Get file → | |
| BF16 | 244.3 GB (5 parts) | 245.6 GB | Too big | Get file → |
For 8 GB VRAM, use IQ1_M (34.22 GB file, ~35.5 GB memory at 8k context). For 16 GB VRAM, Q4_K_S (71.71 GB file, ~73 GB memory) or Q4_K_M (76.54 GB file, ~77.8 GB memory) are options. For 24 GB VRAM, Q5_K_S (86.39 GB file, ~87.7 GB memory) provides higher quality. Memory estimates are approximate.
What GPU do you need for Qwen3.5-122B-A10B?
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 the main file (0.4 GB file, about 1.7 GB in use).
- 6 GB VRAM: the best quant that fits is the main file (0.4 GB file, about 1.7 GB in use).
- 8 GB VRAM: the best quant that fits is the main file (0.4 GB file, about 1.7 GB in use).
- 12 GB VRAM: the best quant that fits is the main file (0.4 GB file, about 1.7 GB in use).
- 16 GB VRAM: the best quant that fits is the main file (0.4 GB file, about 1.7 GB in use).
- 24 GB VRAM: the best quant that fits is the main file (0.4 GB file, about 1.7 GB in use).
- 32 GB VRAM: the best quant that fits is the main file (0.4 GB file, about 1.7 GB in use).
- 48 GB VRAM: the best quant that fits is IQ3_S (46.6 GB file, about 47.9 GB in use).
Which Qwen3.5-122B-A10B quant should you choose?
Q4_K_M is the usual starting point for Qwen3.5-122B-A10B: it balances file size and quality well. Here it is a 76.5 GB file.
The Q8_0 file is 129.9 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 the first one in the table at 0.4 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 Qwen3.5-122B-A10B
How much VRAM does Qwen3.5-122B-A10B need?
About 77.8 GB at Q4_K_M with an 8K context window. That is the 76.5 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 Qwen3.5-122B-A10B on an 8 GB GPU?
Yes, with the smaller files: the 0.4 GB file needs about 1.7 GB at 8K context. Larger quants in this repository do not fit in 8 GB.
What is the file size of Qwen3.5-122B-A10B GGUF?
It depends on the quant: file is 0.4 GB, IQ1_M is 34.2 GB, IQ2_XXS is 36.6 GB, IQ2_M is 39.1 GB, Q2_K_XL is 41.8 GB, IQ3_XXS is 44.7 GB and 18 more.
What context length does Qwen3.5-122B-A10B support?
The GGUF header lists a trained context length of 262,144 tokens (256K). The KV cache grows with the context you use: about 0.8 GB at 8K and 3.2 GB at 32K.
Where can I download Qwen3.5-122B-A10B GGUF?
The files are on Hugging Face in the unsloth/Qwen3.5-122B-A10B-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 Qwen3.5-122B-A10B 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 Qwen3.5‑122B‑A10B on 8 GB VRAM?
Use the IQ1_M quant; it requires about 35.5 GB memory at 8k context, which is beyond 8 GB, so it is not feasible on 8 GB VRAM.
Does this model support vision?
Yes, the model includes a vision encoder and supports vision‑language tasks.
What license applies to this model?
The model is released under the Apache‑2.0 license.