Keye VL 2.0 30B A3B needs about 19.9 GB of memory at Q4_K_M (a 18.6 GB file) with an 8K context window, so it fits on a 24 GB GPU. This repository has 5 quants, from 14.8 GB (Q3_K_M) to 61.3 GB (BF16). Longer context needs more: at 32K tokens the KV cache alone adds about 3.2 GB.
Keye VL 2.0 30B A3B – a 30B multimodal model for long‑video understanding and agent tasks
Keye VL 2.0 30B A3B is a 30‑billion‑parameter multimodal model built on the keye‑vl2 architecture. It supports vision and long‑context reasoning up to 262,144 tokens. The model is released under the Apache‑2.0 license and is available in several GGUF quantizations.
Good at
- Long‑video understanding
- Temporal grounding
- Agent‑ready multimodal tasks
Limits
- No text‑only inference without vision
- Requires large VRAM for high‑precision quant
- No explicit language support listed
Best for
- Video analysis and summarization
- Temporal event detection
- Multimodal agent workflows
Who it’s for. Developers building video‑centric AI applications or agents that need to process long sequences of visual data. The model is suitable for research and production environments that can accommodate its memory demands.
Quants & memory
| Quant | File | Est. memory | vs VRAM | Result | |
|---|---|---|---|---|---|
| Q3_K_M | 14.8 GB | 16.1 GB | Too big | Get file → | |
| Q4_K_M | 18.6 GB | 19.9 GB | Too big | Get file → | |
| Q8_0 | 32.6 GB | 33.9 GB | Too big | Get file → | |
| F16 | 61.3 GB | 62.6 GB | Too big | Get file → | |
| BF16 | 61.3 GB | 62.6 GB | Too big | Get file → |
For 8 GB VRAM, use Q3_K_M (14.76 GB file, ~16.1 GB memory at 8 k context). For 12 GB VRAM, Q4_K_M (18.62 GB file, ~19.9 GB memory). Higher‑precision Q8_0 or F16 requires 32 GB or more VRAM.
What GPU do you need for Keye VL 2.0 30B A3B?
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 Q4_K_M (18.6 GB file, about 19.9 GB in use).
- 32 GB VRAM: the best quant that fits is Q4_K_M (18.6 GB file, about 19.9 GB in use).
- 48 GB VRAM: the best quant that fits is Q8_0 (32.6 GB file, about 33.9 GB in use).
Which Keye VL 2.0 30B A3B quant should you choose?
Q4_K_M is the usual starting point for Keye VL 2.0 30B A3B: it balances file size and quality well. Here it is a 18.6 GB file.
The Q8_0 file is 32.6 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 Q3_K_M at 14.8 GB, with the quality trade-offs that come with fewer bits. A reasonable choice when a 4-bit file will not fit, mostly on larger models. Quality loss is noticeable but often acceptable.
New to quant names? Read GGUF quantization explained.
Specifications
Architecture from the GGUF header
Source
Frequently asked questions about Keye VL 2.0 30B A3B
How much VRAM does Keye VL 2.0 30B A3B need?
About 19.9 GB at Q4_K_M with an 8K context window. That is the 18.6 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 Keye VL 2.0 30B A3B on an 8 GB GPU?
Not at an 8K context. The smallest file needs about 16.1 GB, so you would need a larger GPU or CPU offloading.
What is the file size of Keye VL 2.0 30B A3B GGUF?
It depends on the quant: Q3_K_M is 14.8 GB, Q4_K_M is 18.6 GB, Q8_0 is 32.6 GB, F16 is 61.3 GB, BF16 is 61.3 GB.
What context length does Keye VL 2.0 30B A3B 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 Keye VL 2.0 30B A3B GGUF?
The files are on Hugging Face in the Kwai-Keye/Keye-VL-2.0-30B-A3B-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 Keye VL 2.0 30B A3B 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 Keye VL 2.0 30B A3B on 8 GB VRAM?
Use the Q3_K_M quantization; it needs about 16 GB of VRAM at 8 k context, so 8 GB is insufficient.
Does the model support text‑only inference?
The model is vision‑enabled and designed for multimodal input; text‑only use is not its primary focus.
What license applies to this model?
The model is released under the Apache‑2.0 license.