Mendable

Mendable is an AI answers and enterprise search platform that lets teams ingest their documentation, websites, and knowledge sources and serve grounded, cited AI chat answers over them. Built by the Firecrawl / Mendable team, it exposes a REST API at https://api.mendable.ai/v1 for creating conversations, asking questions over ingested content with streaming, ingesting and managing data sources, and rating answers.

4 APIs 0 Features

AIAnswersEnterprise SearchRAGSupport

APIs

Mendable Chat & Answers API

Ask natural-language questions over ingested documentation and knowledge sources via the mendableChat endpoint, returning grounded answers with citations. Supports Server-Sent E...

Mendable Conversations API

Create a new conversation to obtain a conversation_id used to group a series of chat interactions into a single grounded session.

Mendable Data Ingestion & Sources API

Ingest websites, sitemaps, GitHub repos, Docusaurus sites, YouTube transcripts, single URLs, and raw documents into a Mendable project, and poll ingestion task status. Returns a...

Mendable Ratings API

Submit positive or negative ratings on individual answer messages via rateMessage to feed answer-quality analytics and improvement loops.

Collections

Mendable API

OPEN

Pricing Plans

Mendable Plans Pricing

2 plans

PLANS

Rate Limits

Mendable Rate Limits

4 limits

RATE LIMITS

FinOps

Mendable Finops

FINOPS

Resources

👥

GitHubOrganization

Sources

opencollection: 1.0.0
info:
  name: Mendable API
  version: '1.0'
request:
  auth:
    type: bearer
    token: '{{bearerToken}}'
items:
- info:
    name: Conversations
    type: folder
  items:
  - info:
      name: Create a new conversation
      type: http
    http:
      method: POST
      url: https://api.mendable.ai/v1/newConversation
      body:
        type: json
        data: '{"api_key":"{{apiKey}}"}'
    docs: Creates a new conversation and returns a conversation_id.
- info:
    name: Chat
    type: folder
  items:
  - info:
      name: Ask a question over ingested content
      type: http
    http:
      method: POST
      url: https://api.mendable.ai/v1/mendableChat
      body:
        type: json
        data: '{"api_key":"{{apiKey}}","question":"How do I get started?","history":[],"shouldStream":false}'
    docs: Asks a grounded question over ingested content. Streams Server-Sent Events by default; set shouldStream to false
      for a single JSON response.
- info:
    name: Ingestion
    type: folder
  items:
  - info:
      name: Ingest a data source by URL
      type: http
    http:
      method: POST
      url: https://api.mendable.ai/v1/ingestData
      body:
        type: json
        data: '{"api_key":"{{apiKey}}","url":"https://docs.example.com","type":"website-crawler"}'
    docs: Ingests a website, sitemap, GitHub repo, Docusaurus site, YouTube transcript, or single URL. Returns a task_id.
  - info:
      name: Ingest raw documents
      type: http
    http:
      method: POST
      url: https://api.mendable.ai/v1/ingestDocuments
      body:
        type: json
        data: '{"api_key":"{{apiKey}}","documents":[{"content":"Document text","metadata":{}}]}'
    docs: Ingests up to 500 raw documents (2MB per request). Returns a task_id.
  - info:
      name: Check ingestion task status
      type: http
    http:
      method: POST
      url: https://api.mendable.ai/v1/ingestionStatus
      body:
        type: json
        data: '{"api_key":"{{apiKey}}","task_id":0}'
    docs: Returns the current step, status, and completion metrics for an ingestion task.
- info:
    name: Ratings
    type: folder
  items:
  - info:
      name: Rate an answer message
      type: http
    http:
      method: POST
      url: https://api.mendable.ai/v1/rateMessage
      body:
        type: json
        data: '{"api_key":"{{apiKey}}","message_id":0,"rating_value":1}'
    docs: Submits a positive (1) or negative (-1) rating for an answer message.