rkllm

Automated script to convert Huggingface and GGUF models to rkllm format for running on Rockchip NPU

Qwen3-VL-2B on the RK3588 NPU

OpenAI-compatible API server for Rockchip NPU (RK3588/RK3576) — local LLM inference on Orange Pi with Open WebUI support

Run RKLLM only in one docker command. Already converted models here.

SmolVLM2 on the RK3588 NPU

Entire pipeline for running optimized LLM models on the RK3588S, both on the NPU (RKLLM) or the CPU (llama)

Qwen3-VL-4B on the RK3588 NPU

rkllm rust ffi binding

RK3588 edge AI deployment guides and examples for YOLO11 with RKNN and DeepSeek-R1-Distill-Qwen-1.5B with RKLLM.

Qwen3-VL video on the RK3588 NPU

SmolVLM2 on the RK3588 NPU

A fork of exo-explore/exo that adds RKLLM inference engine support for Rockchip RK3588/RK3576 NPU devices.📱💻 🖥️⌚

InternVL3 VLM on the RK3588 NPU

Terraform provider for Turing Pi 2.5 BMC. Use with freed-dev-llc/modules/turingpi for cluster deployment.

SmolVLM2 on the RK3588 NPU

Pedantic, beginner-friendly setup guide for the Firefly RK182X 3D RAM Stacking Development Kit (AIO-GS1N2 + Core-3588JD4 + RK1820/RK1828) — from unboxing to running an LLM on the NPU.

🌌 Advance multimodal AI with InternVL3.5-4B for RK3588 NPU, enhancing vision and language understanding in a versatile and efficient manner.

Qwen3.5-VL-4B on the RK3588 NPU

Qwen3.5-VL-2B on the RK3588 NPU

Qwen3.5-VL-0.8B on the RK3588 NPU