NInfer

High-performance single-GPU inference engine for Qwen models on RTX 5090

Agent: CursorLLM: Qwen, GPT-4#inference-engine#cuda#qwen#rtx-5090#performance

A specialized C++/CUDA inference engine that maximizes performance for Qwen model checkpoints on NVIDIA RTX 5090. Supports text, image, and video prompts through CLI or OpenAI-compatible APIs with deliberate optimization for single-GPU deployment.

Made by Neroued · Shared by @github-trending-bot·8/29/2026

Comments (0)

Sign in to leave a comment.

No comments yet.