MachineLearningModel & AI Weights Schema (JSON-LD Blueprint)
Structured data for open-weights machine learning models, LLM checkpoints, and ONNX models with parameter counts and GPU memory specs.
Displayed below main page header or above the tool container. • Zero CLS Container
Production JSON-LD Code Block
Embed this structured data block directly into your HTML <head> or framework layout template.
{
"@context": "https://schema.org",
"@type": "SoftwareApplication",
"name": "OmniSearch-7B Dense Retrieval Transformer",
"applicationCategory": "DeveloperApplication",
"operatingSystem": "Linux, macOS (Metal), Windows (CUDA)",
"runtimePlatform": "PyTorch 2.4, ONNX Runtime, Hugging Face Transformers",
"memoryRequirements": "16GB VRAM for FP16 inference, 8GB for INT4 quantized",
"downloadUrl": "https://huggingface.co/omniseotools/omnisearch-7b",
"softwareVersion": "1.2.0",
"codeRepository": "https://github.com/omniseotools/omnisearch-model",
"license": "https://opensource.org/licenses/Apache-2.0",
"description": "7-billion parameter embedding model fine-tuned for semantic SEO taxonomy matching, entity disambiguation, and SERP query intent classification."
}Typed Properties Specification (6 Fields)
Key Schema.org properties and sub-objects required to qualify for rich snippet treatment in Google Search.
| Property | Expected Type | Status | Description & Example |
|---|---|---|---|
| applicationCategory | Text | Required | Classified as 'DeveloperApplication'. e.g. DeveloperApplication |
| runtimePlatform | Text | Required | AI inference runtimes (PyTorch, ONNX, TensorRT, vLLM). e.g. PyTorch 2.4, ONNX Runtime |
| memoryRequirements | Text | Required | VRAM and GPU hardware specifications for FP16/quantized execution. e.g. 16GB VRAM for FP16 inference |
| downloadUrl | URL | Required | Direct model weights repository link (e.g. Hugging Face). e.g. https://huggingface.co/... |
| codeRepository | URL | Required | Model training and inference script repo. e.g. https://github.com/... |
| license | URL | Required | Weights license terms (Apache-2.0, MIT, Llama Community). e.g. https://opensource.org/licenses/Apache-2.0 |
Google Search Console Pitfalls & Critical Warnings
Avoid these 3 common validation traps, parser errors, and algorithmic policy penalties when deploying SoftwareApplication schema.
Missing VRAM / GPU Hardware memoryRequirements
Failing to state GPU memory requirements causes developers with underpowered hardware to face out-of-memory crashes.
Document exact VRAM thresholds (e.g. '16GB VRAM for FP16').
Missing Permissive vs Commercial License Specification
Failing to clarify if commercial use is permitted under model weights license.
Explicitly link Apache 2.0, MIT, or custom commercial license.
Dead Hugging Face downloadUrl
Linking to an unauthenticated or deleted model repository.
Verify public repository availability.
Recommended Implementation Scenarios
Related Schema Blueprints
Explore complementary structured data blueprints across the Technical SEO cluster.
Dataset & Open Data Repository Schema (JSON-LD Blueprint)
SoftwareSourceCode & Git Repository Schema (JSON-LD Blueprint)
Python Package (PyPI) & Wheel Distribution Schema (JSON-LD)
TechWhitepaper & Research Paper Schema (JSON-LD Blueprint)
Displayed below main page header or above the tool container. • Zero CLS Container