Bank data, self-hosted

Add the BankSync MCP server to LibreChat and ask about your transactions, balances and investments from your own self-hosted chat.

Get Started Setup guide

Banks and cards

Total Checking••7731
Savor Card••8830
Savings••5502

Transactions

Oct 14Whole Foods−$86.42
Oct 12Netflix−$15.49
Oct 10Shell−$48.10

LibreChatBankSync MCP

Any MCP client

Another assistant tomorrow?

Same accounts. Whatever you use after LibreChat reads them through the same connection, and no tool moves money.

LibreChat reading Total Checking and Savings through BankSync, with the model you choose.

Connect LibreChat in three steps.

01
Create a scoped key
Connect your banks in BankSync, then create an API key with read scopes under Developers.
New API key
NameMCP client
Scopesbanks:readfeeds:read
Keybsk_••••••••3f9a
02
Add the server
Add BankSync under mcpServers in librechat.yaml and restart. Each user can bring their own key with customUserVars.
librechat.yaml
```
mcpServers:
  banksync:
    type: streamable-http
    url: https://mcp.banksync.io
    headers:
      X-API-Key: "bsk_your_key_here"
```
03
Ask with any model
Pick a model with tool calling and ask about your money.
What did we spend on groceries this month, and is it more than September?
get_transactions · 23 rows · Total Checking ••7731
$642.18 across 23 trips, $58 more than September. Whole Foods is most of it.

What you can ask.

Private financial chat: Query bank data through a self-hosted chat interface where conversations never leave your infrastructure.
Multi-model financial analysis: Compare financial insights from different AI models (GPT-4, Claude, Gemini) using the same bank data.
Team financial dashboard: Deploy a shared LibreChat instance for your team to query financial data with role-based access.
Custom financial assistants: Build custom AI assistants with preset financial queries and workflows in LibreChat.
Spending analysis: Analyze spending patterns, income trends, and account balances through conversational AI.
Budget planning: Create budget plans using any AI model with access to real transaction and balance data.

Change assistants. Keep your data.

BankSync

Spreadsheets

Workspaces

AI you choose

Your code

What LibreChat gets.

Server: Remote and hosted at https://mcp.banksync.io over Streamable HTTP. Nothing to install.
Tools: 227 documented tools: bank data, feeds and jobs, tables, dashboards, automations and enrichments. Tools reference
Sign-in: A scoped API key, sent in the X-API-Key header.
Config: librechat.yaml, under the top-level mcpServers key.
Access: Set by the scopes you grant. Read scopes keep the connection read-only; no tool can move money.
Rate limits: 20 tool calls a minute on Standard, 40 on Professional, 60 on Business, shared across a workspace.
Banks: 14,000+ institutions in the US, Canada, the UK, Europe and Australia, including Consumer Data Right connections.
Guides: LibreChat setup guide Every MCP client MCP server

Questions, answered.

How do I connect LibreChat to BankSync?

Generate a BankSync API key, then add the BankSync MCP server configuration in your LibreChat settings. LibreChat will discover 227 financial data tools for querying bank data.

Can I self-host LibreChat with BankSync?

Yes. LibreChat is fully open-source and self-hostable. BankSync MCP works with self-hosted instances, giving you complete control over your AI chat interface while accessing live bank data.

What AI models work with LibreChat and BankSync?

LibreChat supports multiple AI providers including OpenAI, Anthropic, Google, and local models. Any model with tool-calling support can use BankSync MCP tools for financial queries.

Is my financial data private with self-hosted LibreChat?

With self-hosted LibreChat, your chat history and conversations stay on your infrastructure. BankSync data is fetched over encrypted connections with scoped API keys.

Self-hosted financial AI

Get Started →LibreChat setup guide