Nexa SDK runs any model on any device, across any backend locally—text, vision, audio, speech, or image generation—on NPU, GPU, or CPU. It supports Qualcomm, Intel, AMD and Apple NPUs, GGUF, Apple MLX, and the latest SOTA models (Gemma3n, PaddleOCR).
Free
Launch Team / Built With
CloudflareStill paying to deploy your apps? Try Cloudflare Workers for free
Promoted
This is a game-changer. "Run, build & ship local AI in minutes" is exactly the promise developers have been waiting for. If Nexa SDK delivers on simplifying the local AI toolchain, it will unlock a new wave of privacy-focused and low-latency applications. Incredibly excited to try this out!
Congrats on launch Alex! This AI tool is exactly what the industry needs right now.
@audrey_adams thank you for your support, we are working on more developer features
@audrey_adams Thanks Audrey! Local AI is private, cost-efficient, and always available. It is the future of on-device AI infra.
@ray_luan thanks!
@ray_luan Looking forward to hear your feedback
Super techy & strong team! Must try!
@rachel_hu huge thanks for u!
@rachel_hu Thanks for the support!!!
Really amazing new release team!
Exciting to launch this car
Well product