Grida Standard Library is a 100% free & open-source handpicked collection of design assets used by our Editor and our AI.

• fonts
• shapes
• symbols
• logos
• patterns
• wallpapers
• icons
• colors
• sizes
• objects
• components
• styles
• illustrations
• 3d-illustrations
• doodles
• scribbles

{
  "object": "o",
  "id": "o_x1234",
  "name": "logo-grida.svg",
  "description": "Grida logo svg with black path fill",
  "type": "application/svg+xml",
  "keywords": ["logo", "grida"],
  "category": "logo",
  "categories": ["logos"],
  "author": {
    "name": "Grida",
    "url": "https://grida.co/grida"
  },
  "version": 1,
  "width": 64,
  "height": 64,
  "bytes": 1234,
  "url": "https://grida.co/library/v1/o_x1234",
  "urls": {
    "raw": "https://grida.co/library/v1/o_x1234"
  },
  "transparency": true,
  "visual_padding": [2, 2, 2, 2],
  "color": "#000000",
  "colors": ["#000000"],
  "background": "#FFFFFF",
  "svg": "<svg>...</svg>",
  "score": 0.5,
  "year": 2025,
  "created_at": 1677610602,
  "updated_at": 1677610602
}

• generator - the generator used to create the asset

  • grida-canvas
  • the-bundle
  • dall-e-3
  • midjourney
  • photoshop

• entropy - the visual complexity of the design

  • 0 ~ 1

Setup

uv venv
source .venv/bin/activate
uv pip install -r requirements.txt

unsplash.py: Download images from Unsplash.

python photos/unsplash.py xR4Yt3AEXLY --download --q=regular --access-key="..." --dir=/path/to/out

optimize.py: Optimize images.

python photos/optimize.py ~/Public/library/category  ~/Public/library/category/out --max-size=3

metadata.py: Generate metadata for images.

python photos/metadata.py /path/to/process

outputs .metadata.json files.

describe.py: Describe images using ollama.

Needs ollama installed. Needs model with vision support.

python describe.py /path/to/process --model=gemma3:27b

outputs .describe.json files.

object.py: Combines all metadata for library uploads.

python object.py /path/to/process

outputs .object.json files.

upload.py: Upserts the object (file) and metadata (row) to the Grida Library.

python upload.py /path/to/process