serper-search-mcp-server
Related Servers
Alternatives to serper-search-mcp-server
No user-submitted related servers found.
Related Servers
- AlicenseAqualityDmaintenanceWeb search for AI agents across 6 engines (Serper, Brave, Exa, Tavily, Firecrawl, Perplexity) through one search tool. Routes each query to the cheapest engine that clears a quality bar and caches repeats. Hosted, streamable-HTTP, BYOK supported.11MIT
- AlicenseCqualityBmaintenanceOne endpoint, five search providers. Search broker for AI agents with automatic fallback, RRF ranking, and budget enforcement. The LiteLLM of web search.1555 PyPI9MIT
- FlicenseNot gradedqualityDmaintenanceA universal web search and content retrieval tool for AI agents, supporting multiple search providers and content extraction.-
- AlicenseAqualityFmaintenanceFree multi-source web search server for AI agents, with confidence scoring and token optimization.3Apache 2.0
- AlicenseNot gradedqualityDmaintenanceSelf-hosted retrieval router for AI agents: one MCP/REST endpoint that routes web search across Brave, Tavily, Exa, Firecrawl, and Crossref (BYOK) with per-request cost ceilings, deadlines, and an auditable route receipt on every response.Apache 2.0
- AlicenseNot gradedqualityDmaintenanceUnified web layer for AI agents. Search across 8 engines, stealth browse with Camoufox/Playwright, manage cookie auth, and act on 24 platforms.28MIT
TDQS
Scored across 15 tools
The vertical search tools (search_web, search_images, search_news, search_scholar, etc.) are clearly distinct from each other, but the unified `search` tool explicitly subsumes nine of them, so an agent must guess between `search(type=web)` and `search_web`. The web-retrieval family is also crowded — search, search_web, search_many, search_rag_context, and deep_research all return web results with only subtle intent differences (raw vs chunked vs synthesized vs parallel).
Most tools follow a predictable `search_<vertical>` or verb_noun pattern (search_web, search_news, search_rag_context, resolve_links, deep_research). The lone outlier is `autocomplete` (no verb) and the bare `search`, which breaks the prefix family while sitting in the middle of it.
Fifteen tools is within a reasonable band, but roughly nine of them are redundant duplicates of the unified `search` tool, so the set is effectively two overlapping surfaces shipped together rather than 15 distinct capabilities. The redundancy inflates the count without adding reachable functionality.
Coverage of the search domain is strong: all major Google verticals, scraping, link resolution, RAG-ready chunking, and synthesized research are present. Minor gaps remain around pagination/result-set continuation, batch scraping, and any cache or credit-budget inspection tool, but agents can work around these.