BigML MCP Server
Overview
BigML MCP Server is an MCP-compatible integration that connects MCP clients to BigML’s machine learning platform. It lets you build, train, and operationalize machine learning models through standardized MCP calls using a single static server URL.
- Type: MCP server / developer tool
- Category: AI integration (MCP servers)
- Provider: BigML via Pipedream
- Server URL:
https://mcp.pipedream.net/v2
Features
MCP Integration
- Provides a static MCP server URL usable across compatible clients.
- Authentication handled when adding the server to your MCP-enabled application.
- Designed to work with multiple chat or MCP clients (configured per client).
The server exposes BigML operations as MCP tools (currently 3 actions):
-
Create Source (Remote URL)
- Create a BigML data source from a remote URL.
- BigML downloads the data file directly from the provided URL.
- Intended as the first step in the ML workflow (ingesting data into BigML).
-
Create Model
- Create a machine learning model in BigML.
- Can be based on:
- A source ID
- A dataset ID
- An existing model ID
- Supports building supervised models as part of an end-to-end ML pipeline.
-
Create Batch Prediction
- Generate batch predictions using an existing supervised model and dataset.
- Requires:
- A Supervised Model ID
- A Dataset ID
- Enables operationalizing trained models at scale via batch inference.
Operationalization & Deployment
- Enables company-wide operationalization of machine learning using BigML through MCP.
- Can run in any environment where your MCP client operates, while BigML handles the ML workload in the cloud or on-prem (per BigML’s platform capabilities).
Pricing
Pricing information is not provided in the available content. Use of this MCP server may be subject to:
- BigML account / plan pricing, and
- Any applicable Pipedream or platform usage terms.
For current pricing, refer to BigML and/or Pipedream pricing documentation directly.
