QORA-4B is a 4-billion parameter language model with built-in vision. Used Burn

[blockmandev]

Pure Rust multimodal inference engine based on Qwen3.5-4B. No Python, no CUDA, no external ML frameworks. Single executable + model weights = portable AI that runs on any machine.

GPU accelerated — auto-detects Vulkan (Windows/Linux) or Metal (macOS) GPU and runs inference on it. Falls back to CPU if no GPU available. Smart system awareness — detects RAM and CPU at startup and adjusts generation limits automatically.

Try: https://huggingface.co/qoranet/QORA-LLM-4B