Airtable Mcp Server

Connect AI assistants to your Airtable bases using the Model Context Protocol (MCP). This actor enables Claude, ChatGPT, and other AI tools to read, write, search, and manage your Airtable data through 15 powerful operations.

Why Use This Actor?

• AI-Ready: Works with Claude Desktop, ChatGPT, and any MCP-compatible AI assistant
• Complete Coverage: All 15 Airtable operations - from simple reads to schema management
• Two Modes: Run as a persistent MCP server (Standby) or execute one-time operations (Batch)
• No Code Required: Just provide your Airtable API key and start interacting
• Secure: Your API key is never stored - used only during execution

Available Operations (15 MCP Tools)

Data Operations

Tool	Description
list_records	Retrieve records with optional filtering using Airtable formulas
search_records	Full-text search across all fields in a table
get_record	Fetch a single record by its ID
create_record	Add new records to any table
update_records	Modify existing records (single or batch)
delete_records	Remove records (single or batch)

Schema Management

Tool	Description
list_bases	List all Airtable bases you have access to
list_tables	Get all tables within a specific base
describe_table	Get detailed table schema including all fields
create_table	Create new tables with custom fields
update_table	Modify table name and description
create_field	Add new fields to existing tables
update_field	Modify field properties

Collaboration

Tool	Description
create_comment	Add comments to records
list_comments	Retrieve all comments on a record

Quick Start

1. Get Your Airtable API Key

1. Go to Airtable Token Creation
2. Create a Personal Access Token with these scopes:
   • schema.bases:read - List bases and tables
   • data.records:read - Read records
   • data.records:write - Create/update/delete records
   • schema.bases:write - Create tables and fields (optional)

2. Run in Batch Mode

Execute a single operation and get results:

{
    "airtableApiKey": "patXXXXXXXX.XXXXXXXX",
    "mode": "batch",
    "batchOperation": "list_bases"
}

3. Run in Standby Mode (MCP Server)

For AI assistant integration, run with mode: "standby" and enable Actor Standby in settings.

Usage Examples

List All Your Bases

{
    "airtableApiKey": "patXXXXXXXX.XXXXXXXX",
    "mode": "batch",
    "batchOperation": "list_bases"
}

Get Records with Filter

{
    "airtableApiKey": "patXXXXXXXX.XXXXXXXX",
    "mode": "batch",
    "batchOperation": "list_records",
    "baseId": "appXXXXXXXXXXXX",
    "tableId": "Tasks",
    "filterFormula": "{Status}='Active'",
    "maxRecords": 50
}

Search Across All Fields

{
    "airtableApiKey": "patXXXXXXXX.XXXXXXXX",
    "mode": "batch",
    "batchOperation": "search_records",
    "baseId": "appXXXXXXXXXXXX",
    "tableId": "Contacts",
    "searchTerm": "john@example.com"
}

Create a New Record

{
    "airtableApiKey": "patXXXXXXXX.XXXXXXXX",
    "mode": "batch",
    "batchOperation": "create_record",
    "baseId": "appXXXXXXXXXXXX",
    "tableId": "Tasks",
    "recordData": {
        "Name": "New Task",
        "Status": "Todo",
        "Priority": "High",
        "Due Date": "2026-01-15"
    }
}

Create a New Table

{
    "airtableApiKey": "patXXXXXXXX.XXXXXXXX",
    "mode": "batch",
    "batchOperation": "create_table",
    "baseId": "appXXXXXXXXXXXX",
    "tableName": "Projects",
    "tableDescription": "Track all company projects",
    "fields": [
        {"name": "Name", "type": "singleLineText"},
        {"name": "Status", "type": "singleSelect", "options": {"choices": [{"name": "Planning"}, {"name": "Active"}, {"name": "Completed"}]}},
        {"name": "Due Date", "type": "date"}
    ]
}

Integrating with Claude Desktop

To use this actor as an MCP server with Claude Desktop:

1. Enable Standby Mode in Actor Settings
2. Add to Claude Desktop config (claude_desktop_config.json):

{
    "mcpServers": {
        "airtable": {
            "url": "https://airtable-mcp-server.apify.actor/mcp?token=YOUR_APIFY_TOKEN",
            "headers": {
                "Authorization": "Bearer YOUR_AIRTABLE_TOKEN"
            }
        }
    }
}

3. Restart Claude Desktop - Airtable tools will be available

Input Reference

Parameter	Type	Required	Description
airtableApiKey	string	Yes	Your Airtable Personal Access Token
mode	string	No	standby (default) or batch
batchOperation	string	No	Operation to perform (see tools above)
baseId	string	No	Base ID (e.g., appXXXXXXXXXX)
tableId	string	No	Table ID or name
recordId	string	No	Single record ID
recordIds	array	No	Multiple record IDs for batch operations
recordData	object	No	Field values for create/update
filterFormula	string	No	Airtable formula for filtering
searchTerm	string	No	Text for full-text search
maxRecords	integer	No	Maximum records to return (default: 100)
tableName	string	No	Name for new/updated table
tableDescription	string	No	Description for table
fieldName	string	No	Name for new field
fieldType	string	No	Type for new field
fieldOptions	object	No	Options for field (e.g., select choices)
comment	string	No	Comment text

Output

Results are stored in the default dataset with the following structure:

{
    "operation": "list_records",
    "id": "recXXXXXXXXXX",
    "createdTime": "2026-01-02T10:00:00.000Z",
    "fields": {
        "Name": "Example Record",
        "Status": "Active"
    }
}

Use Cases

• AI-Powered Data Entry: Let Claude or ChatGPT add records to your Airtable
• Natural Language Queries: Ask AI to find specific records using plain English
• Automated Reporting: Generate reports from Airtable data via AI
• Schema Management: Create and modify tables through conversational AI
• Workflow Automation: Integrate Airtable operations into Apify workflows

Pricing

This actor uses Apify platform resources. Typical costs:

• Batch operations: ~$0.001 per run
• Standby mode: Billed per compute unit while active

Attribution

This actor wraps airtable-mcp-server by Adam Jones (domdomegg), licensed under MIT.

Support

• Airtable API Documentation
• MCP Protocol Specification
• Report Issues

License

MIT License - see LICENSE for details.