Read chat threads and message history, then send replies through the real 4based messaging endpoints. Wire your own model, your own scripts and your own routing on top of clean JSON, without a residential proxy or a password in the loop.
CreatorAPI gives you the 4based messaging surface: list chats, read history, and send with an optional price for a paid unlock. What runs on top is yours to design, whether that is an AI assistant for your chatters, a canned reply engine, or a routing layer that hands hard conversations to a human. You keep control of tone and policy, we keep the connection stable.
List conversations with chats, then read full message history per thread, so your tool always has the context a good reply needs.
Send replies through the same native endpoint the app uses, with message_price set for a paid unlock.
Feed history to Claude, ChatGPT or your own model and post the reply back. The API stays neutral about what you run.
A signed messages.received event fires the moment a fan writes, so automations react instantly instead of polling.
4based needs no residential proxy, unlike OnlyFans and Fansly, so you connect a session and start reading and sending right away.
Read unread counts per chat, mark one thread or every thread as received, and pin a conversation so your queue reflects reality.
Read or set a custom message price for one specific chat, on top of the account's default rate.
Open a chat with a fan who does not have one yet, instead of only replying to inbound.
Either way your 4based password never touches CreatorAPI unless you explicitly opt into encrypted automatic reconnect.
# list a creator's own id, then a chat thread's messages curl https://api.creator-api.com/v1/{account}/native/user/{selfId}/chat/{chatId}/message \ -H "X-API-Key: ca_live_your_key"
No. It gives you native read and send access. The logic, the model and the tone are yours, which keeps you in control of policy and quality.
Yes. Read the thread through the API, generate a reply with any model, and send it back. A native MCP server also lets an agent do this directly.
No. 4based needs no residential proxy, so chatting automation runs off a signed in session without extra infrastructure.
Yes. The same read and send pattern covers OnlyFans and Fansly natively under one key.
Yes. Unread counts, marking one or all chats as received, and pinning a conversation are all native calls, so your own inbox view can match what a human agent would see.
Get your key, connect a 4based account and read your first thread in minutes.