Ling 3.0 Flash VL needs about 91.3 GB of memory at Q4_K_M (a 75.3 GB file) with an 8K context window, which is more than a 48 GB GPU holds. This repository has 23 quants, from 0.5 GB to 88.3 GB (Q5_K_M). Longer context needs more: at 32K tokens the KV cache alone adds about 62.0 GB.
Ling 3.0 Flash VL – 124 B multimodal model with vision and long‑context support
Ling 3.0 Flash VL is a 124 B multimodal model that integrates image and video understanding into text reasoning. It supports a context window of up to 256 K tokens and activates only 5.5 B parameters per token. The model is released under the MIT license.
Good at
- Vision‑enabled reasoning
- Long‑context multimodal dialogue
- Agentic task planning
Limits
- Requires large VRAM for high‑quality quants
- No native audio support
- Limited to 256 K token context
Best for
- Image‑based question answering
- Video summarisation
- Multimodal agent workflows
Who it’s for. Developers building multimodal applications that need both vision and extensive context. The model is suitable for research and production use where a large context window and efficient parameter activation are required.
Quants & memory
| Quant | File | Est. memory | vs VRAM | Result | |
|---|---|---|---|---|---|
| — | 0.5 GB | 16.5 GB | Too big | Get file → | |
| IQ1_S | 25.2 GB | 41.2 GB | Too big | Get file → | |
| IQ1_M | 28.0 GB | 44.0 GB | Too big | Get file → | |
| IQ2_XXS | 32.7 GB | 48.7 GB | Too big | Get file → | |
| IQ2_XS | 36.5 GB | 52.5 GB | Too big | Get file → | |
| IQ2_S | 36.8 GB | 52.8 GB | Too big | Get file → | |
| IQ2_M | 40.6 GB | 56.6 GB | Too big | Get file → | |
| Q2_K_S | 41.9 GB | 58.0 GB | Too big | Get file → | |
| Q2_K | 45.4 GB | 61.4 GB | Too big | Get file → | |
| IQ3_XXS | 47.9 GB | 63.9 GB | Too big | Get file → | |
| IQ3_XS | 50.8 GB | 66.9 GB | Too big | Get file → | |
| IQ3_S | 53.8 GB | 69.8 GB | Too big | Get file → | |
| Q3_K_S | 53.8 GB | 69.8 GB | Too big | Get file → | |
| IQ3_M | 54.3 GB | 70.3 GB | Too big | Get file → | |
| Q3_K_M | 59.3 GB | 75.3 GB | Too big | Get file → | |
| Q3_K_L | 64.4 GB | 80.4 GB | Too big | Get file → | |
| IQ4_XS | 66.4 GB | 82.4 GB | Too big | Get file → | |
| Q4_0 | 70.5 GB | 86.5 GB | Too big | Get file → | |
| Q4_K_S | 70.7 GB | 86.7 GB | Too big | Get file → | |
| Q4_K_M | 75.3 GB | 91.3 GB | Too big | Get file → | |
| Q4_1 | 78.0 GB | 94.0 GB | Too big | Get file → | |
| Q5_K_S | 85.8 GB | 101.8 GB | Too big | Get file → | |
| Q5_K_M | 88.3 GB | 104.4 GB | Too big | Get file → |
For 8 GB VRAM, use the IQ1_S quant (25.2 GB file, ~41 GB memory at 8 k context). For 16 GB VRAM, IQ2_XS (36.53 GB file, ~52.5 GB memory) is a good balance. Higher‑quality quants such as Q4_K_M or Q5_K_M require 32 GB or more VRAM.
What GPU do you need for Ling 3.0 Flash VL?
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: no quant fits at an 8K context.
- 24 GB VRAM: the best quant that fits is the main file (0.5 GB file, about 16.5 GB in use).
- 32 GB VRAM: the best quant that fits is the main file (0.5 GB file, about 16.5 GB in use).
- 48 GB VRAM: the best quant that fits is IQ1_M (28.0 GB file, about 44.0 GB in use).
Which Ling 3.0 Flash VL quant should you choose?
Q4_K_M is the usual starting point for Ling 3.0 Flash VL: it balances file size and quality well. Here it is a 75.3 GB file.
If memory is tight, the smallest file is the first one in the table at 0.5 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 Ling 3.0 Flash VL
How much VRAM does Ling 3.0 Flash VL need?
About 91.3 GB at Q4_K_M with an 8K context window. That is the 75.3 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 Ling 3.0 Flash VL on an 8 GB GPU?
Not at an 8K context. The smallest file needs about 16.5 GB, so you would need a larger GPU or CPU offloading.
What is the file size of Ling 3.0 Flash VL GGUF?
It depends on the quant: file is 0.5 GB, IQ1_S is 25.2 GB, IQ1_M is 28.0 GB, IQ2_XXS is 32.7 GB, IQ2_XS is 36.5 GB, IQ2_S is 36.8 GB and 17 more.
What context length does Ling 3.0 Flash VL support?
The GGUF header lists a trained context length of 131,072 tokens (128K). The KV cache grows with the context you use: about 15.5 GB at 8K and 62.0 GB at 32K.
Where can I download Ling 3.0 Flash VL GGUF?
The files are on Hugging Face in the mradermacher/Ling-3.0-flash-VL-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 Ling 3.0 Flash VL 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 Ling 3.0 Flash VL on 8 GB VRAM?
Only the smallest quants such as IQ1_S can fit, but memory usage will be close to the VRAM limit.
Does the model support video input?
Yes, it includes a ViT visual encoder and VideoRoPE for spatial and temporal encoding.
What license applies to this model?
The model is released under the MIT license.