Free Live Data MCP
Related Servers
Alternatives to Free Live Data MCP
No user-submitted related servers found.
Related Servers
- AlicenseNot gradedqualityDmaintenanceProvides live data API for AI agents with 51 REST endpoints across finance, travel, weather, news, and more, using USDC micropayments without API keys.14 npmMIT

ReefAPI MCPofficial
AlicenseAqualityBmaintenanceOne MCP server providing access to 160+ live web data APIs (search, social media, e-commerce, real estate, jobs, travel, news, finance, and more) using dynamic discovery via 4 generic tools to avoid the agent's tool limit.54MIT
Stophy MCPofficial
AlicenseNot gradedqualityBmaintenanceEnables AI agents to discover and query public web data across web search, YouTube, Reddit, TikTok, Instagram, LinkedIn, maps, shopping, jobs, real estate, and finance through three tools for finding endpoints, inspecting their inputs and costs, and executing calls that return markdown or JSON.74 npmMIT- AlicenseNot gradedqualityBmaintenanceEnables AI agents to discover official public data sources, inspect their metadata and endpoint parameters, and execute live queries through a unified response envelope with a 500 ms budget.MIT

BatchData MCP Serverofficial
AlicenseNot gradedqualityBmaintenanceEnables AI assistants and agents to query U.S. real estate data, including property lookups, comparables, property searches, skip tracing, phone/address verification, and investor buy boxes, through plain-language requests.1MIT- AlicenseNot gradedqualityCmaintenanceOfficial AnyAPI agent skills — hundreds of scraping & data APIs through one gateway: one key, USD pay-per-request, normalized schemas, automatic failover.1Apache 2.0
TDQS
Scored across 4 tools
The tools map to a clean discovery→inspect→execute pipeline, with search_apis (sources) and search_endpoints (operations) targeting different resource levels. The only mild overlap is that an agent might start with search_endpoints when it should first enumerate sources, but the descriptions clarify the workflow.
All four names follow a verb_noun snake_case pattern (search_apis, search_endpoints, describe_endpoint, call_endpoint). The verb choice also encodes the workflow stage (search → describe → call), which is predictable.
Four tools is lean but proportionate to a read-only live-data catalog: discovery, inspection, and execution each earn their place. It sits just below the ideal mid-range breadth but nothing feels redundant.
The surface covers the full read lifecycle: find an API, find an operation, learn its parameters, then call it. Gaps are minor — there is no tool to verify enrollment/API-key status or check remaining quota before consuming it, though errors from call_endpoint likely surface these.