Hi HN, I built KonbiniAPI because every social media scraping API I tried returned raw platform JSON with different field names, structures, and conventions per platform. You end up writing adapters and fixing breaking changes instead of building your product.

KonbiniAPI normalizes everything into ActivityStreams 2.0 (the same W3C standard behind the Fediverse). A TikTok user and an Instagram user come back with the same fields: type, preferredUsername, followerCount, etc. Build once, add new platforms without code changes.

What it does:

- 30 endpoints across Instagram (10) and TikTok (20)

- Profiles, posts, comments, stories, search, followers, transcripts, and more

- Every request fetches live public data — nothing cached

- No rate limits, simple credit-based pricing

- One request = one credit, failed requests are refunded

Tech: Hono + Chanfana, OpenAPI schema auto-generated from endpoint definitions, and in-house scraping infrastructure.

Site: https://konbiniapi.com

Docs: https://docs.konbiniapi.com

OpenAPI spec: https://docs.konbiniapi.com/openapi.json

Happy to answer any questions about the architecture, the normalization approach, or the scraping challenges.