https://github.com/aep/moxgo

cross platform inference runtime
https://github.com/aep/moxgo

Last synced: about 1 month ago
JSON representation

cross platform inference runtime

• Host: GitHub
• URL: https://github.com/aep/moxgo
• Owner: aep
• License: apache-2.0
• Created: 2026-05-22T09:03:47.000Z (3 months ago)
• Default Branch: main
• Last Pushed: 2026-06-02T07:48:24.000Z (2 months ago)
• Last Synced: 2026-06-28T08:22:40.931Z (about 1 month ago)
• Language: C
• Homepage:
• Size: 165 KB
• Stars: 3
• Watchers: 0
• Forks: 0
• Open Issues: 0
• Metadata Files:
  • Readme: README.md

Awesome Lists containing this project

README

          
MOXGO - the ollama for ONNX

============================

get an easy http api for ONNX models on android,windows,linux,mac,ios with native acceleration.

    > moxgo serve

the builtin cli makes it easy to play with popular classification models

    > moxgo pull birdnet

    > moxgo run birdnet --file what_bird_is_this.mp3

    > moxgo pull yolo11n

    > moxgo run yolo11n --file is_this_a_dog.png

adding custom models

--------------------

you can add your own ONNX models without publishing them to the registry.

place the model files in `~/.moxgo/models//` with a `manifest.json` and a `model.onnx`.

    ~/.moxgo/models/my-model/

    ├── manifest.json

    ├── model.onnx

    └── labels.csv        (optional)

the manifest describes inputs, outputs, and runtime options:

```json

{

  "inputs": {

    "images": {

      "type": "image",

      "width": 640,

      "height": 640

    }

  },

  "outputs": {

    "output0": {

      "labels": "coco80"

    }

  }

}

```