New

Akorbit Desktop is now available — automate directly from your desktop on macOS & Windows.

Download
Akorbit API

Production-ready APIs
for your sovereign AI.

Integrate RAG, Text-to-SQL and AI agents into your applications in a few lines of code. Secure REST APIs, webhooks, official SDKs.

Bearer Auth REST / JSON HTTPS/TLS Webhooks
View documentation Get an API key
POST /api/v1/agents/{id}/query
# Query an AI agent on your HR data

curl -X POST \
  "https://platform.akorbit.com/api/v1/agents/ag_rh_001/query" \
  -H "Authorization: Bearer <YOUR_TOKEN>" \
  -H "Content-Type: application/json" \
  -d '{
    "query": "What is the remote work policy?",
    "user_id": "user_482",
    "stream": false
  }'

# API response
{
  "answer": "The remote work policy allows 3 days/week...",
  "sources": ["Convention_RH_2024.pdf"],
  "confidence": 0.97
}
Authentication

Secured by Bearer token

All API requests are authenticated via a Bearer token. Manage your keys from the Akorbit console with granular permission levels.

1
Create your account

Sign up at platform.akorbit.com and access your developer workspace.

2
Generate an API key

Create an API key with the permissions suited to your use case.

3
Integrate in a few lines

Use our SDKs or call the REST API directly from your application.

Authorization header
Authorization: Bearer ak_live_••••••••••••••••••••••••••••••••
Permission levels
Read read
Query collections and retrieve answers
Ingest ingest
Upload and index documents and vectors
Agents agents
Create, configure and run AI agents
Admin admin
Manage users, collections and settings
Endpoints

Key endpoints

A consistent and documented API for all your needs: document ingestion, RAG, AI agents, Text-to-SQL and workflows.

POST /documents/ingest Ingestion Ingest and index a document
Request body (multipart/form-data)
file         → rapport_Q3.pdf
collection_id→ "col_finance_001"
ocr          → true
language     → "fr"
Response
{
  "document_id": "doc_8f2a...",
  "status": "indexed",
  "chunks": 142,
  "ocr_applied": true
}
POST /vectors/ingest Ingestion Ingest vector embeddings
Request body (JSON)
{
  "collection_id": "col_rh_001",
  "vectors": [{
    "id": "chunk_001",
    "embedding": [0.12, 0.87, ...],
    "text": "Article 3 — Contract term: 12 months",
    "metadata": {"source": "doc.pdf"}
  }]
}
Response
{
  "inserted": 48,
  "collection_id": "col_rh_001",
  "status": "success"
}
POST /chat/query RAG Query a collection with RAG
Request body
{
  "collection_id": "col_finance_001",
  "query": "What is the Q3 net margin?",
  "top_k": 5,
  "stream": false
}
Response
{
  "answer": "The Q3 net margin is 18.4%...",
  "sources": [{
    "file": "Rapport_Q3.pdf",
    "page": 12,
    "score": 0.94
  }]
}
POST /agents/{agent_id}/query Agents Run an AI agent on a query
Request body
{
  "query": "What is my leave policy?",
  "user_id": "user_482",
  "session_id": "sess_abc123",
  "stream": false
}
Response
{
  "answer": "According to the 2024 HR agreement, you are entitled to...",
  "agent_id": "ag_rh_001",
  "sources": ["Convention_RH.pdf"],
  "confidence": 0.97
}
POST /sql/query Text-to-SQL Convert a question to SQL and execute
Request body
{
  "datasource_id": "db_erp_001",
  "query": "Top 5 products by revenue this quarter?",
  "execute": true
}
Response
{
  "sql": "SELECT name, SUM(amount)...",
  "results": [{"name":"Alice",...}],
  "rows": 5
}
POST /workflows/{workflow_id}/run Workflow Trigger an automated workflow
Request body
{
  "input": {
    "document_url": "https://...",
    "action": "summarize_and_tag"
  }
}
Response
{
  "run_id": "run_7d9f...",
  "status": "running",
  "webhook_url": "https://..."
}
Use cases

What you can build

Document chatbot

Embed an AI assistant into your intranet or internal tools.

  • Sourced answers from your documents
  • Multi-language support
  • Conversation history

Business AI agent

Create specialized agents for HR, finance, support or sales.

  • Access to your internal systems
  • Automated actions triggered
  • Human escalation when needed

Document pipeline

Automate extraction, classification and indexing of documents.

  • Automatic OCR
  • Category-based classification
  • Real-time notifications

Workflow automation

Trigger complex workflows from your business events.

  • Webhook triggers
  • Sequential processing chains
  • Logs and monitoring

Data analysis

Connect your data warehouses and query them in natural language.

  • SQL generated automatically
  • Results in structured JSON
  • Compatible with all SQL warehouses

Product integration

Embed Akorbit AI directly into your SaaS product.

  • White-labeling available
  • Per-user rate limiting
  • Built-in analytics dashboard
SDKs

Integrate in your favorite language

Python
Official SDK
JavaScript
Node.js / Browser
PHP
Official SDK
REST / cURL
Compatible with all languages
Access the full documentation

Ready to integrate AI into your product?

Get your API key and start building in minutes.

Get an API key View documentation
API key available instantly
Interactive documentation
Python, JS, PHP SDKs
Dedicated developer support