Instagram Post Image Analyzer

See what is actually in your Instagram images, at scale. Point it at a profile or a list of posts and a vision model describes every image, detects objects and brands, reads on-image text, flags brand safety, and suggests hashtags.

What you get

For every post:

• Post basics: shortcode, URL, media type, caption, like and comment counts, thumbnail URL, author
• Description: one sentence describing the image
• Category: the content category
• Objects: notable objects detected in the image
• Brands: visible brands or logos
• On-image text: any text read from the image
• Brand safety: safe, review, or risky
• Suggested hashtags: relevant hashtags for the image

Use cases

• Auto tag and categorize a creator's or competitor's feed
• Detect product placements and brand logos across posts
• Build searchable image catalogs from Instagram content
• Brand safety screening before a partnership
• Generate hashtag ideas from real post imagery

How to use

1. Enter a Username to analyze recent posts, or paste specific Post URLs
2. Set Max posts when using a username
3. Optionally change the Vision model
4. Run the actor. Each post becomes one row in the Dataset tab

Output example

{
  "shortcode": "DZ8cY6xAElr",
  "postUrl": "https://www.instagram.com/p/DZ8cY6xAElr/",
  "mediaType": "video",
  "category": "wildlife",
  "description": "A close-up of a baby sea turtle swimming through clear turquoise water above a sandy seabed.",
  "objects": ["baby sea turtle", "water", "seabed"],
  "brands": [],
  "onImageText": [],
  "brandSafety": "safe",
  "suggestedHashtags": ["BabySeaTurtle", "OceanLife", "WildlifePhotography", "MarineLife", "SeaTurtle"]
}

Notes

• For videos and Reels, the cover image is analyzed.
• Posts without an available image are returned with empty analysis fields so no post is silently dropped.
• The default model is a fast, low cost vision model. Any vision model that accepts image input can be used.