Youtu VL 4B Instruct needs about 6.4 GB of memory at Q8_0 (a 5.2 GB file) with an 8K context window, so it fits on a 8 GB GPU. This repository has 2 quants, from 5.2 GB (Q8_0) to 9.8 GB (F16). Longer context needs more: at 32K tokens the KV cache alone adds about 2.9 GB.
Youtu VL 4B Instruct – a compact vision‑language model with 4.9B parameters
Youtu VL 4B Instruct is a 4.9B parameter Vision‑Language Model built on the deepseek2 architecture. It supports vision tasks without task‑specific modules and uses Vision‑Language Unified Autoregressive Supervision. The model is available in GGUF format.
Good at
- Vision‑centric tasks such as classification and segmentation
- General multimodal reasoning and VQA
- OCR and multi‑image understanding
Limits
- No vision support beyond standard VLM tasks
- License is other, not specified
- Requires 8k context for full performance
Best for
- Image classification on edge devices
- Multimodal chat with visual input
- Rapid prototyping of vision‑language applications
Who it’s for. Developers building lightweight vision‑language applications who need a small footprint model. Researchers testing VLUAS concepts on a 4B parameter scale.
Quants & memory
| Quant | File | Est. memory | vs VRAM | Result | |
|---|---|---|---|---|---|
| Q8_0 | 5.2 GB | 6.4 GB | Fits | Get file → | |
| F16 | 9.8 GB | 11.0 GB | Too big | Get file → |
For 8 GB VRAM, use the Q8_0 quantization (file size 5.21 GB, estimated memory 6.4 GB at 8k context). If more VRAM is available, F16 (9.8 GB, 11 GB memory) offers higher precision. Memory estimates are approximate.
What GPU do you need for Youtu VL 4B Instruct?
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 Q8_0 (5.2 GB file, about 6.4 GB in use).
- 12 GB VRAM: the best quant that fits is F16 (9.8 GB file, about 11.0 GB in use).
Which Youtu VL 4B Instruct quant should you choose?
Q8_0 (5.2 GB) is the middle option in this repository. Close to lossless. It costs roughly twice the memory of a 4-bit file, so it suits cases where you have plenty of VRAM.
New to quant names? Read GGUF quantization explained.
Specifications
Architecture from the GGUF header
Source
Frequently asked questions about Youtu VL 4B Instruct
How much VRAM does Youtu VL 4B Instruct need?
About 6.4 GB at Q8_0 with an 8K context window. That is the 5.2 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 Youtu VL 4B Instruct on an 8 GB GPU?
Yes, with Q8_0: the 5.2 GB file needs about 6.4 GB at 8K context. Larger quants in this repository do not fit in 8 GB.
What is the file size of Youtu VL 4B Instruct GGUF?
It depends on the quant: Q8_0 is 5.2 GB, F16 is 9.8 GB.
What context length does Youtu VL 4B Instruct support?
The GGUF header lists a trained context length of 32,768 tokens (32K). The KV cache grows with the context you use: about 0.7 GB at 8K and 2.9 GB at 32K.
Where can I download Youtu VL 4B Instruct GGUF?
The files are on Hugging Face in the tencent/Youtu-VL-4B-Instruct-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 Youtu VL 4B Instruct use?
Hugging Face lists the license as other. 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 Youtu VL 4B Instruct on 8 GB VRAM?
Yes, the Q8_0 quantization requires about 6.4 GB of VRAM at 8k context, fitting within an 8 GB budget.
Does this model support vision tasks?
Yes, it is a Vision‑Language Model with comprehensive vision‑centric capabilities such as classification, detection, and segmentation.
What license does the model use?
The model is released under an "other" license, not specified in the provided data.