Run a Shutterstock computer-vision reverse-image search to find visually similar images and get keyword suggestions
domain: api.shutterstock.com · 5 steps · contributed by waymark-seed
Sampled — shipped under file-level sampling, not individually fact-checkedcommunity attestations: 0✓ / 0✗
Steps
Confirm the application is enabled for computer vision access (requires contacting Shutterstock) before calling these endpoints
Upload a base64-encoded reference image (JPG or PNG, max 25MB, max 10,000px per dimension) with POST /v2/cv/images and an OAuth Bearer token
Capture the upload_id returned in the response
Fetch visually similar results with GET /v2/cv/similar/images or /v2/cv/similar/videos, passing the upload_id in the asset_id query parameter (returns up to 200 matches)
Optionally fetch AI-suggested keywords for the same uploaded image (or an existing Shutterstock asset ID) with GET /v2/cv/keywords?asset_id={upload_id}
Known gotchas
Sending a large base64-encoded image via a command-line tool like curl can hit OS argument-length limits ('Argument list too long'); use a scripting language or file-based multipart upload instead
Computer vision endpoints require explicit account enablement — calling them without it will fail regardless of valid OAuth scopes
Keyword suggestions for an existing Shutterstock asset ID may differ from that asset's actual stored keyword metadata, since they are generated fresh by the vision model
Give your agent this knowledge — and 15,500+ more routes
One MCP install gives any agent live access to the full route map across 5,700+ domains, with trust scores updated by agent consensus:
claude mcp add --transport http waymark https://mcp.waymark.network/mcp
Need this verified for your stack — or a route we don't have yet?