The Meridian Edge MCP server lets AI assistants access real-time prediction market data directly — no manual API calls required. Ask Claude about today's NBA consensus and get a live answer.
Model Context Protocol (MCP) is a standard for connecting AI assistants to external data sources. Instead of requiring developers to build custom integrations, MCP servers expose tools that LLMs can call directly during a conversation.
The Meridian Edge MCP server wraps our consensus API into a set of named tools — enabling any MCP-compatible client (Claude Desktop, Cursor, etc.) to query prediction market data in natural language.
get_consensus — fetch current consensus for a sport or eventget_opportunities — surface high-divergence events (Starter+)get_account — check your API key usage and remaining callsAdd this to your claude_desktop_config.json:
{{
"mcpServers": {{
"meridian-edge": {{
"command": "npx",
"args": ["-y", "@meridianedge/mcp-server"],
"env": {{
"MERIDIAN_API_KEY": "me_live_your_key_here"
}}
}}
}}
}}Once configured, ask Claude:
The MCP server uses the same API key as direct REST access. Get an API key (100 calls/day) at meridianedge.iothe pricing page.
Get an API key and add it to your AI assistant in minutes.
View Plans →Need more? Starter ($29/mo) — 1,000 calls/day + divergence signals