Run a lookalike search
SetINSTAGRAM_USERNAME to a known Instagram handle, then run this with the TypeScript SDK:
Read the result
An abbreviated, fictional result:similarity.score ranges from 0 to 1. Review it alongside shared_traits. To evaluate the new creators against a specific brief, continue with campaign scoring. To retrieve linked accounts, use each returned creator ID with creator lookup.
Choose and weight your seeds
A seed is a reference creator. Provide 1–10 seeds, identified by creator ID or Instagram username. You can mix these formats:404 seed_not_found. Check the handle or choose another seed before submitting another request.
Narrow the results
Add filters when your roster needs specific account sizes or engagement levels:Load another page
Whenhas_more is true, repeat the request with the same seeds and filters, adding the returned next_cursor as cursor. Treat the cursor as an opaque string. Stop when has_more is false.
Each delivered page is billed per returned creator. See Pagination for the request pattern and the Lookalike reference for all fields.
Run a complete roster workflow
Find up to five similar creators from one to three seed handles, then score them against a campaign brief. The complete program makes at most two requests and costs up to 12.5 credits ($0.125) with API-key billing. Download the MIT-licensed workflow source or browse it on GitHub. Requires Node.js 22+ and pnpm 10.32.1.pnpm type-check and pnpm test for synthetic checks. Review actual returned creators and charges separately with your own key. A failure stops the program; follow Error Handling before submitting again. Browse the cookbook for other workflows.