Leni AI
Send your API key on every call. Conversations go through POST /users/run-models. The API may also call /quick-reply to enrich follow-ups when you supply questionHistory. Use GET /users/sdk/me for the key profile, and the /users/me/chat-* routes to read persisted chat data (same key auth).
Endpoints in this section
| Method | Path | Purpose |
|---|---|---|
| POST | /users/run-models | Prompts, attachments, tagging, summaries, sessions |
| GET | /users/sdk/me | Profile for the key (optional extra payload) |
| GET | /users/me/chat-messages | List chat messages (newest first) |
| GET | /users/me/chat-messages/:messageId | One message row (owner) |
| GET | /users/me/chat-sessions/:sessionId | Session metadata when the owner participates |
| GET | /users/me/chat-sessions/:sessionId/messages | Transcript for the session |
| POST | /users/me/chat-sessions/attachments | Resolve file anchors in summaries (Leni /resolve-file-url) |
| POST | /users/me/chat-sessions/message-responses | Raw message_response rows from the database |
Guides
- Run models — request fields, files, examples.
- Quick reply —
questionHistoryand how it differs from the main request. - Summarize — screen capture, URL, and short-summary fields.
- Sessions —
sessionId. - Chat messages — list and fetch messages; poll after a run-models deadline response.
- Chat sessions — session row and full transcript.
- Resolve attachments and Message responses — scoped POST bodies shared with
ChatSessionScopedBodyDto.
Memory APIs are under Memory; custom analysts under Analysts.