Context engineering for agentic and enterprise workflows

One API connects email, documents, tools, and the web into answers ready for reasoning and automation, without building RAG.

Go to playground Talk to an expert

Trusted by the world’s most innovative developers

How it works

The fastest path to real context

Immediate integration

SDK and live playground to test requests and responses quickly.

Connect datasources

iGPT deeply indexes email, docs, and work data automatically.

One unified endpoint

Send a single request that runs retrieval, context shaping, and reasoning in one pipeline.

Contextual responses

Responses are ready for reasoning and automation, with citations to the source.

Ship agents

No parsing, chunking, indexing, or prompt tuning to maintain.

Why iGPT

Build with context in a single unified call

curl -X POST https://api.igpt.ai/v1/recall/ask/ \
    -H "Content-Type: application/json" \
    -H "Authorization: Bearer $IGPT_API_KEY" \
    -d '{
        "user": "user_123",
        "input": "Open action items from yesterday’s board meeting",
        "stream": false,
        "quality": "cef-1-normal",
        "output_format": { 
            "schema": {
                ...
            } 
        }
    }'

from igpt import IGPTClient
from schemas.action_items import ACTION_ITEMS_SCHEMA
import os

client = IGPTClient(
    api_key=os.environ["IGPT_API_KEY"],
    user="user_123"
)

response = client.ask(
    input="Open action items from yesterday’s board meeting",
    stream=False,
    quality="cef-1-normal",
    output_format=ACTION_ITEMS_SCHEMA
)

print(response)

{
  "schema": {
    "type": "object",
    "properties": {
      "action_items": {
        "type": "array",
        "description": "List of action items",
        "items": {
          "type": "object",
          "properties": {
            "title": {
              "type": "string",
              "description": "Short summary of the action item"
            },
            "owner": {
              "type": "string",
              "description": "Person responsible for the action item"
            },
            "due_date": {
              "type": "string",
              "format": "date",
              "description": "Expected completion date"
            }
          },
          "required": [
            "title",
            "owner",
            "due_date"
          ],
          "additionalProperties": false
        }
      }
    },
    "required": ["action_items"],
    "additionalProperties": false
  }
}

Input

No parsing, no vector stores, no prompt chains. Just call the API with your preferred language.

Read docs

Response.json

{
    "action_items": [
        {
        "title": "Approve revised Q1 budget allocation",
        "owner": "Dvir Ben-Aroya",
        "due_date": "2026-01-15",
      },
      {
        "title": "Approve final FY2026 strategic priorities",
        "owner": "Board of Directors",
        "due_date": "2026-01-31",
      },
      ...
    ]
}