KAITO

KAITO (Kubernetes AI Toolchain Operator) is an open-source operator suite that automates LLM model inference, fine-tuning, and Retrieval Augmented Generation (RAG) engine deployment in Kubernetes clusters. It simplifies the process of deploying large AI models through optimized preset configurations and integrates with Karpenter for GPU node auto-provisioning.

1 APIs 0 Features

AIGPUInferenceKubernetesLLMMachine LearningOpen SourceOperatorRAG

APIs

KAITO RAGEngine API

RAGEngine exposes endpoints for managing retrieval-augmented generation services with embedded vector databases, including document indexing, retrieval, and chat completion endp...

Collections

KAITO RAGEngine REST API

OPEN

Pricing Plans

Kaito Plans Pricing

3 plans

PLANS

Rate Limits

Kaito Rate Limits

5 limits

RATE LIMITS

FinOps

Kaito Finops

FINOPS

Resources

GitHubOrganization

Sources

opencollection: 1.0.0
info:
  name: KAITO RAGEngine REST API
  version: 1.0.0
items:
- info:
    name: Index Management
    type: folder
  items:
  - info:
      name: Create or append documents to an index
      type: http
    http:
      method: POST
      url: http://localhost:8080/index
      body:
        type: json
        data: '{}'
    docs: Create a RAG index (or append to it) with the supplied documents.
  - info:
      name: List documents in an index
      type: http
    http:
      method: GET
      url: http://localhost:8080/indexes/:index_name/documents
      params:
      - name: index_name
        value: ''
        type: path
      - name: limit
        value: ''
        type: query
      - name: offset
        value: ''
        type: query
      - name: max_text_length
        value: ''
        type: query
      - name: metadata_filter
        value: ''
        type: query
    docs: Return documents stored in the named index with optional pagination and filters.
  - info:
      name: Append documents to an index
      type: http
    http:
      method: POST
      url: http://localhost:8080/indexes/:index_name/documents
      params:
      - name: index_name
        value: ''
        type: path
      body:
        type: json
        data: '{}'
    docs: Append additional documents (each with a doc_id) to an existing index.
  - info:
      name: Delete documents from an index
      type: http
    http:
      method: POST
      url: http://localhost:8080/indexes/:index_name/documents/delete
      params:
      - name: index_name
        value: ''
        type: path
      body:
        type: json
        data: '{}'
    docs: Delete documents from the named index by doc_id.
  - info:
      name: Delete an index
      type: http
    http:
      method: DELETE
      url: http://localhost:8080/indexes/:index_name
      params:
      - name: index_name
        value: ''
        type: path
    docs: Delete the entire named index.
- info:
    name: Persistence
    type: folder
  items:
  - info:
      name: Persist an index to storage
      type: http
    http:
      method: POST
      url: http://localhost:8080/persist/:index_name
      params:
      - name: index_name
        value: ''
        type: path
      body:
        type: json
        data: '{}'
    docs: Persist the named index to storage at an optional path.
  - info:
      name: Load an index from storage
      type: http
    http:
      method: POST
      url: http://localhost:8080/load/:index_name
      params:
      - name: index_name
        value: ''
        type: path
      body:
        type: json
        data: '{}'
    docs: Load a persisted index back into the RAGEngine.
- info:
    name: Retrieval
    type: folder
  items:
  - info:
      name: Retrieve nodes from a RAG index
      type: http
    http:
      method: POST
      url: http://localhost:8080/retrieve
      body:
        type: json
        data: '{}'
    docs: Retrieve the top relevant nodes from a RAG index for a given query.
- info:
    name: Chat
    type: folder
  items:
  - info:
      name: RAG-grounded chat completions
      type: http
    http:
      method: POST
      url: http://localhost:8080/v1/chat/completions
      body:
        type: json
        data: '{}'
    docs: OpenAI-compatible chat completion endpoint that grounds responses in the specified RAG index.
bundled: true