24 GB accelerator memory
Q4 GGUF · llama.cpp
It loads, but context and concurrency should be reduced because Q4 weights consume most of the card.
Qwen
Text model
A fast multimodal Qwen model with fewer refusal behaviors.
Capabilities
The most useful reasons to choose this model, without making you read through its repository first.
Open-ended assistants
Tool-using agents
Visual question answering
For providers
The provider requirements live here, separate from the information you need to choose and use the model.
24 GB accelerator memory
Q4 GGUF · llama.cpp
It loads, but context and concurrency should be reduced because Q4 weights consume most of the card.
48 GB accelerator memory
Q4 GGUF · llama.cpp
The 21–23 GB Q4 weights need room for context, vision, and concurrent requests.
Source and license
OpenMayhem links back to the source repository so you can inspect the model card, files, license, limitations, and creator guidance directly.
Uncensored edition by HauHau