Local by default.
Inference runs on your hardware through the default localhost API. Online downloads and updates are distinct from local model execution.
ABOUT NINFEREZ
Make local AI
feel approachable.
NInferEZ makes NInfer simpler to set up and use on Windows, bringing model management and everyday inference controls into a familiar workspace.
Built by 2beng2.
Built on open work.
Manager is the front door to the NInferEZ ecosystem. It connects desktop workflows with independent Engine and Converter tools, so you can begin simply and explore deeper when you need to.
NInferEZ Engine is an independent distribution of ninfer-all, itself based on NInfer. The ecosystem retains upstream authorship, licenses, and third-party notices. Our model artifacts convert and package existing weights; original model authorship remains credited.
By
THE WAY WE BUILD
Practical. Local. Transparent.
Clarity over guesswork.
Managed profiles, compatibility checks, and understandable memory estimates help you make the next choice.
Credit where it belongs.
Model provenance, upstream acknowledgments, and component licenses are part of the distribution, not an afterthought.
Follow the project.
Explore Manager on GitHub, report issues, and find NInfer model conversions on Hugging Face.
YOUR NEXT STEP
Put your GPU to work.
Your models, your workspace, your next idea.
Start with NInferEZ Manager.
Windows x64 · Installer or Portable
ACKNOWLEDGMENTS
The work behind
the workspace.
NInferEZ makes NInfer easier to set up and use on Windows. The original runtime and model work belong to their authors.
NInfer
The original inference runtime, created by Neroued and its contributors. Model execution and CUDA kernels come from this upstream engine.
ninfer-all
The Windows distribution maintained by iamwavecut, which is the base of NInferEZ Engine. Its inherited NInfer-3090 work and contributor history are retained in the upstream records.
NInferEZ · 2beng2
The Manager interface, guided workflows, model organization, Windows packaging, and documented compatibility and integration additions. See our change record for the scope of the Engine distribution.
Models & components
Qwen supplies the base model family. Swift, OrcaRouter, Heretic, and their direct checkpoint and quantization sources retain their original authorship. Draft components and bundled dependencies keep their own credits and licenses.
Original sources in our model collection →