Spark MCP 2.0 adds an intelligence layer on top of GWI’s data, with new tools, guardrails and guided workflows. The result is more consistent, reliable answers in the AI tool you already work in.
More power under the hood - New tools and guardrails help the assistant find the right data and give consistent answers.
Scope first - Market and timeframe are locked in before any answer is pulled, so results land on the right market and period.
Audience definitions stay fixed - Pick a saved audience or describe a new one, and it’s applied to every answer, with no drift between follow-up questions.
Guided workflows - Built-in steps take the assistant from question to answer.
How to migrate to Spark MCP 2.0
Switching to Spark MCP 2.0 takes a few minutes. Workspace admins set up the new connector first, then each team member connects to it.
1. IT or Development teams (Workspace admins)
Disable the current GWI Spark (v1) connector.
Create a new connector using the updated URL - (https://api.globalwebindex.com/v2/spark-api/mcp).
Find step-by-step instructions for your LLM in our Spark MCP collection:
2. Workspace users/members
Go to “Connectors,” search for “GWI Spark” and click Connect.
Log in on the GWI login page to enable the connector. You’ll be taken back to your host app automatically.
Open a new chat and go to “Connectors.” Confirm that "GWI Spark" is enabled.