bu-30b-a3b-preview needs about 20.2 GB of memory at Q4_K_M (a 18.9 GB file) with an 8K context window, so it fits on a 24 GB GPU. This repository has 20 quants, from 0.1 GB to 61.1 GB (BF16). Longer context needs more: at 32K tokens the KV cache alone adds about 3.2 GB.
30.5B‑parameter vision‑language model with MoE architecture
bu‑30b‑a3b‑preview is a 30.5B‑parameter Qwen3‑VL model. It supports vision tasks and has a 262144‑token context length. The model is available in multiple GGUF quantizations.
Good at
- Vision‑language reasoning
- Long‑context multimodal tasks
- Browser‑use integration
Limits
- No explicit license info
- Requires large VRAM for higher‑bit quantizations
- No text‑only performance data
Best for
- Image captioning with context
- Video summarization
- Interactive browsing agents
Who it’s for. Developers building multimodal applications that need vision and text understanding on a single GPU. The model is suited for research and production use where browser‑use capabilities are required.
Quants & memory
| Quant | File | Est. memory | vs VRAM | Result | |
|---|---|---|---|---|---|
| — | 0.1 GB | 1.4 GB | Fits | Get file → | |
| IQ1_S | 9.1 GB | 10.4 GB | Too big | Get file → | |
| IQ2_XXS | 9.8 GB | 11.1 GB | Too big | Get file → | |
| IQ1_M | 9.9 GB | 11.2 GB | Too big | Get file → | |
| IQ2_XS | 10.9 GB | 12.2 GB | Too big | Get file → | |
| IQ2_S | 10.9 GB | 12.2 GB | Too big | Get file → | |
| IQ3_XXS | 13.7 GB | 15.0 GB | Too big | Get file → | |
| IQ3_S | 15.3 GB | 16.6 GB | Too big | Get file → | |
| IQ4_XS | 16.5 GB | 17.8 GB | Too big | Get file → | |
| Q4_0 | 17.2 GB | 18.5 GB | Too big | Get file → | |
| IQ4_NL | 17.2 GB | 18.5 GB | Too big | Get file → | |
| Q4_K_S | 18.1 GB | 19.4 GB | Too big | Get file → | |
| Q4_K_M | 18.9 GB | 20.2 GB | Too big | Get file → | |
| Q4_1 | 19.1 GB | 20.4 GB | Too big | Get file → | |
| Q5_0 | 21.0 GB | 22.3 GB | Too big | Get file → | |
| Q5_1 | 22.9 GB | 24.3 GB | Too big | Get file → | |
| Q8_0 | 32.5 GB | 33.8 GB | Too big | Get file → | |
| BF16 | 38.7 GB | 40.0 GB | Too big | Get file → | |
| F16 | 38.7 GB | 40.0 GB | Too big | Get file → | |
| BF16 | 61.1 GB (2 parts) | 62.4 GB | Too big | Get file → |
For 8 GB VRAM, use IQ1_S (9.09 GB file, 10.4 GB memory at 8K context). For 12 GB VRAM, IQ2_S (10.93 GB file, 12.2 GB memory). For 16 GB VRAM, Q4_K_M (18.88 GB file, 20.2 GB memory). These estimates are approximate.
What GPU do you need for bu-30b-a3b-preview?
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 (0.1 GB file, about 1.4 GB in use).
- 6 GB VRAM: the best quant that fits is the main file (0.1 GB file, about 1.4 GB in use).
- 8 GB VRAM: the best quant that fits is the main file (0.1 GB file, about 1.4 GB in use).
- 12 GB VRAM: the best quant that fits is IQ1_M (9.9 GB file, about 11.2 GB in use).
- 16 GB VRAM: the best quant that fits is IQ3_XXS (13.7 GB file, about 15.0 GB in use).
- 24 GB VRAM: the best quant that fits is Q5_0 (21.0 GB file, about 22.3 GB in use).
- 32 GB VRAM: the best quant that fits is Q5_1 (22.9 GB file, about 24.3 GB in use).
- 48 GB VRAM: the best quant that fits is BF16 (38.7 GB file, about 40.0 GB in use).
Which bu-30b-a3b-preview quant should you choose?
Q4_K_M is the usual starting point for bu-30b-a3b-preview: it balances file size and quality well. Here it is a 18.9 GB file.
The Q8_0 file is 32.5 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 0.1 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 bu-30b-a3b-preview
How much VRAM does bu-30b-a3b-preview need?
About 20.2 GB at Q4_K_M with an 8K context window. That is the 18.9 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 bu-30b-a3b-preview on an 8 GB GPU?
Yes, with the smaller files: the 0.1 GB file needs about 1.4 GB at 8K context. Larger quants in this repository do not fit in 8 GB.
What is the file size of bu-30b-a3b-preview GGUF?
It depends on the quant: file is 0.1 GB, IQ1_S is 9.1 GB, IQ2_XXS is 9.8 GB, IQ1_M is 9.9 GB, IQ2_XS is 10.9 GB, IQ2_S is 10.9 GB and 14 more.
What context length does bu-30b-a3b-preview 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 bu-30b-a3b-preview GGUF?
The files are on Hugging Face in the Mungert/bu-30b-a3b-preview-GGUF repository. The download buttons on this page open the file on Hugging Face. ggufmodels.com does not host any model files.
Can I run this model on 8 GB VRAM?
Yes, using the IQ1_S quantization which requires about 10.4 GB of VRAM at 8K context.
Does the model support text‑only inference?
The model is primarily vision‑language; text‑only performance is not highlighted.
What is the context length of this model?
The trained context length is 262144 tokens.