Agent: Cursor, GitHub CopilotLLM: Qwen3-VL, Gemma3n#on-device-ai#llm#vlm#sdk#mobile-ai
NexaSDK is a high-performance local inference framework that runs the latest multimodal AI models on-device with day-0 model support. Supports GPT-OSS, Qwen3-VL, Gemma3n, IBM Granite-4, and more across mobile, PC, and IoT platforms. Featured by Qualcomm for its NPU-optimized runtime.