Chift for AI

Supercharge your AI apps with a secure integration layer

Connect your AI agents to 100+ software with one secure integration. Let them act and retrieve data, safely and at scale.

Trusted by Tech pioneers

Get started

Faster development and unified documentation

For Users

Leverage the full power of our Unified APIs for your AI applications.

Thanks to our AI technology (e.g. MCP), you can push and retrieve data from a lot of financial software in one-click.

Chift is the leading API for AI:

Easy to connect Chift in your AI applications (LLM-providers, AI agents, ...).

Enterprise-grade security.

We are experts on accounting, invoicing, POS, ... use cases.

Unified data giving you a 360° view of your business.

A bunch of different types of logos on a white background.

For developers

The Chift Model Context Protocol (MCP) defines a set of tools you can use to build AI Agents that can interact with the different integrations offered by Chift.

LangChain
Python
docs

from langchain_mcp_adapters.client import MultiServerMCPClient

token = requests.post(
   f"{chift_url}/mcp-token",
   json={
       "CHIFT_CLIENT_SECRET": "your_client_secret",
       "CHIFT_CLIENT_ID": "your_client_id",
       "CHIFT_ACCOUNT_ID": "your_account_id",
       "CHIFT_CONSUMER_ID": "your_consumer_id",
   },
).json()["access_token"]

client = MultiServerMCPClient(
   {
       "Chift MCP Server": {
           "transport": "streamable_http",
           "url": "https://mcp.chift.eu",
           "headers": {"Authorization": f"Bearer {token}"},
       }
   }
)tools = await client.get_tools()

PydanticAI
Python
docs

from pydantic_ai.agent import Agent
from pydantic_ai.mcp import MCPServerStreamableHTTP

mcp_url = "https://mcp.chift.eu"
chift_url = "https://api.chift.eu"

token = requests.post(
   f"{chift_url}/mcp-token",
   json={
       "CHIFT_CLIENT_SECRET": "your_client_secret",
       "CHIFT_CLIENT_ID": "your_client_id",
       "CHIFT_ACCOUNT_ID": "your_account_id",
       "CHIFT_CONSUMER_ID": "your_consumer_id",
   },
).json()["access_token"]

mcp = MCPServerStreamableHTTP(url=mcp_url, headers={"Authorization": f"Bearer {token}"})
agent = Agent("openai:gpt-4o", toolsets=[mcp])

AI SDK
Typescript
docs

import { experimental_createMCPClient } from "ai";
import { StreamableHTTPClientTransport } from "@modelcontextprotocol/sdk/client/streamableHttp.js";

const response = await fetch("https://api.chift.eu/mcp-token", {
 method: "POST",
 headers: {
   "Content-Type": "application/json",
 },
 body: JSON.stringify({
   clientId: "your_client_secret",
   clientSecret: "your_client_id",
   accountId: "your_account_id",
   consumerId: "your_consumer_id",
 }),
});

const token = (await response.json()).access_token;

const httpTransport = new StreamableHTTPClientTransport(
 new URL("https://mcp.chift.eu"),
 {
   requestInit: {
     headers: {
       Authorization: `Bearer ${token}`,
     },
   },
 }
);
const client = await experimental_createMCPClient({
 transport: httpTransport,
});

FastMCP
Python
docs

from fastmcp import Client
from fastmcp.client.transports import StreamableHttpTransport

mcp_url = "https://mcp.chift.eu"
chift_url = "https://api.chift.eu"

token = requests.post(
   f"{chift_url}/mcp-token",
   json={
       "CHIFT_CLIENT_SECRET": "your_client_secret",
       "CHIFT_CLIENT_ID": "your_client_id",
       "CHIFT_ACCOUNT_ID": "your_account_id",
       "CHIFT_CONSUMER_ID": "your_consumer_id",
   },
).json()["access_token"]

client = Client(
   StreamableHttpTransport(
       url=mcp_url,
       headers={"Authorization": f"Bearer {token}"},
   )
)

tools = await client.list_tools()

For Users

Leverage the full power of our Unified APIs for your AI applications.

Thanks to our AI technology (e.g. MCP), you can push and retrieve data from a lot of financial software in one-click.

Chift is the leading API for AI:

Easy to connect Chift in your AI applications (LLM-providers, AI agents, ...).

Enterprise-grade security.

We are experts on accounting, invoicing, POS, ... use cases.

Unified data giving you a 360° view of your business.

A bunch of different types of logos on a white background.

For developers

The Chift Model Context Protocol (MCP) defines a set of tools you can use to build AI Agents that can interact with the different integrations offered by Chift.

LangChain
Python
docs

from langchain_mcp_adapters.client import MultiServerMCPClient

token = requests.post(
   f"{chift_url}/mcp-token",
   json={
       "CHIFT_CLIENT_SECRET": "your_client_secret",
       "CHIFT_CLIENT_ID": "your_client_id",
       "CHIFT_ACCOUNT_ID": "your_account_id",
       "CHIFT_CONSUMER_ID": "your_consumer_id",
   },
).json()["access_token"]

client = MultiServerMCPClient(
   {
       "Chift MCP Server": {
           "transport": "streamable_http",
           "url": "https://mcp.chift.eu",
           "headers": {"Authorization": f"Bearer {token}"},
       }
   }
)tools = await client.get_tools()

PydanticAI
Python
docs

from pydantic_ai.agent import Agent
from pydantic_ai.mcp import MCPServerStreamableHTTP

mcp_url = "https://mcp.chift.eu"
chift_url = "https://api.chift.eu"

token = requests.post(
   f"{chift_url}/mcp-token",
   json={
       "CHIFT_CLIENT_SECRET": "your_client_secret",
       "CHIFT_CLIENT_ID": "your_client_id",
       "CHIFT_ACCOUNT_ID": "your_account_id",
       "CHIFT_CONSUMER_ID": "your_consumer_id",
   },
).json()["access_token"]

mcp = MCPServerStreamableHTTP(url=mcp_url, headers={"Authorization": f"Bearer {token}"})
agent = Agent("openai:gpt-4o", toolsets=[mcp])

AI SDK
Typescript
docs

import { experimental_createMCPClient } from "ai";
import { StreamableHTTPClientTransport } from "@modelcontextprotocol/sdk/client/streamableHttp.js";

const response = await fetch("https://api.chift.eu/mcp-token", {
 method: "POST",
 headers: {
   "Content-Type": "application/json",
 },
 body: JSON.stringify({
   clientId: "your_client_secret",
   clientSecret: "your_client_id",
   accountId: "your_account_id",
   consumerId: "your_consumer_id",
 }),
});

const token = (await response.json()).access_token;

const httpTransport = new StreamableHTTPClientTransport(
 new URL("https://mcp.chift.eu"),
 {
   requestInit: {
     headers: {
       Authorization: `Bearer ${token}`,
     },
   },
 }
);
const client = await experimental_createMCPClient({
 transport: httpTransport,
});

FastMCP
Python
docs

from fastmcp import Client
from fastmcp.client.transports import StreamableHttpTransport

mcp_url = "https://mcp.chift.eu"
chift_url = "https://api.chift.eu"

token = requests.post(
   f"{chift_url}/mcp-token",
   json={
       "CHIFT_CLIENT_SECRET": "your_client_secret",
       "CHIFT_CLIENT_ID": "your_client_id",
       "CHIFT_ACCOUNT_ID": "your_account_id",
       "CHIFT_CONSUMER_ID": "your_consumer_id",
   },
).json()["access_token"]

client = Client(
   StreamableHttpTransport(
       url=mcp_url,
       headers={"Authorization": f"Bearer {token}"},
   )
)

tools = await client.list_tools()

Features

Purpose-built for AI

Ready to power AI solutions

Thanks to our MCP/RAG server, your AI product can easily access financial software and automatically execute a wide range of operations.

A computer screen with a bunch of icons on it.A computer screen with a bunch of icons on it.
A blue and white background with different icons.

Authentication made easy

Authentication has never been easier. Chift automatically handles all authentication methods and securely stores your users’ credentials, ensuring both security and ease of use. By streamlining every tool call, it frees agents to focus on taking action without interruption, delivering a seamless experience from start to finish.

visual with a logo icon ISO 27001 in the middle

Secure by design

Chift ensures strict access control, permissions syncing, and customer-level data isolation. We follow the industry's highest standards of security and privacy and are proud to be ISO 27001 certified.

Take the integration fast lane

Chift is the one-click solution to all your integration needs.
Book a demo