Search api
Your real-time SERP API solution. Mastering proxy management, CAPTCHAs, and JSON parsing for seamless web data extraction.
Verdict
Common use cases
- Validate SEO rankings across multiple engines
- Pull competitor product listings from Amazon
- Research support solutions from YouTube tutorials
- Track brand mentions in real-time news
- Compare pricing across shopping engines
Integration
- Vendor
- Search api
- Category
- developer-tools
- Auth
- API_KEY
- Tools
- 5
- Composio slug
search_api
Tools
- Get Account Info
Tool to retrieve account usage statistics including monthly usage, remaining credits, hourly limits, and subscription period information. Use when you need to check current account status or usage limits.
- Get Cached Search by ID
Tool to retrieve cached search results by search ID in JSON format. Use when you need to access previously executed search results without re-running the query. The search ID is found in the search_metadata.id or search_metadata.json_url fi
- Get Cached Search HTML by ID
Tool to retrieve cached search results by search ID in HTML format. Use when you need to access the raw HTML snapshot of a previous search execution. The search ID is available in the search_metadata.id or search_metadata.html_url field fro
- Get Locations
Tool to get available locations for geo-localized search queries. Use when you need to find location identifiers for targeting Google searches to specific geographic areas. Returns location metadata including identifiers, names, coordinates
- Search
Tool to perform unified search across 40+ search engines including Google, Bing, Yahoo, DuckDuckGo, YouTube, Amazon, and more. Use when you need to retrieve search results with rich structured data including organic results, ads, knowledge
Setup
Setup guide
- 11. Sign up at searchapi.io and copy your API key from the dashboard. 2. In Switchy, open Settings → Integrations → Developer Tools and click 'Connect Search API'. 3. Paste your API key when prompted and click 'Authorize'. 4. Switchy confirms the connection by fetching your account info — you'll see remaining credits and monthly limit. 5. Open any Space and type '@Search api show me top 10 Google results for project management software' to test. 6. The MCP returns structured JSON with titles, URLs, snippets, and metadata. 7. Use the 'Get Account Info' tool anytime to check credit balance. 8. For geo-targeted queries, invoke 'Get Locations' first to find the correct location identifier, then pass it in your search prompt. 9. Cached results persist by search ID — reference them later without burning credits.
What teammates see: by default, memories from Search api are scoped to the Space (PROJECT visibility) - you can mark any memory PRIVATE or share it ORG-wide.
Works well with
Top models
Compatibility data appears once enough Spaces have used this MCP together with a given model.
How Switchy teams use it
Starter prompts
Check SEO Position
@Search api search Google for 'cloud storage solutions' and tell me which position our site appears in the top 50 resultsOpen in a Space →
Compare Shopping Results
@Search api search Amazon and eBay for 'wireless noise cancelling headphones under $200' and show me the top 5 from eachOpen in a Space →
Monitor Brand Mentions
@Search api search Google News for mentions of 'Acme Corp' in the past week and summarize the top 3 storiesOpen in a Space →
Research Video Tutorials
@Search api search YouTube for 'how to export data from Salesforce to CSV' and give me the 3 most viewed videos with linksOpen in a Space →
Audit Credit Usage
@Search api get my account info and tell me how many credits I have left this month and when they resetOpen in a Space →
Example outputs
Illustrative - representative of the model's voice and quality, not literal recordings.
@searchapi find the top 5 Python web scraping libraries on GitHub and summarize their star counts and main features
Based on a GitHub search, here are the top 5 Python web scraping libraries: Beautiful Soup (58.2k stars) — HTML/XML parser with simple navigation; Scrapy (52.1k stars) — full framework with spider architecture and pipeline processing; Selenium (30.4k stars) — browser automation for JavaScript-heavy sites; Requests-HTML (13.8k stars) — combines Requests with parsing and JavaScript rendering; lxml (2.6k stars) — fast XML/HTML processing with XPath support. Beautiful Soup remains most popular for beginners, while Scrapy dominates production environments.
This example demonstrates cross-engine search (GitHub) with AI synthesis. The MCP returns structured search results; the AI extracts and summarizes key metadata. Requires API credits per search; check your account tier's hourly limits before bulk queries.
@searchapi check my remaining credits and tell me if I have enough for 20 more searches today
Your Search API account shows 847 credits remaining this month (resets Jan 1) with a 100-searches-per-hour limit. You've used 12 searches in the current hour. You have sufficient credits for 20 searches, and your hourly rate limit allows it — go ahead. Note that complex queries (with extra parameters like location targeting or HTML caching) may consume additional credits per search.
This read-only example shows account monitoring before heavy usage. The Get Account Info tool returns usage stats; the AI interprets whether your budget supports the planned workload. Useful for teams managing shared API keys across multiple Spaces.
@searchapi search Google for 'best project management tools 2024' from a New York location and cache the HTML results for later review
Search executed with location set to New York, NY (location_id: 1023191). Top results include Monday.com, Asana, ClickUp, Notion, and Jira — all appearing in sponsored and organic positions. HTML snapshot cached with search_id: 'a3f8e2b9'. You can retrieve the full rendered page later using this ID without re-running the query or consuming additional credits. Organic results emphasize collaboration features and AI integrations as 2024 trends.
This example combines geo-targeting, search execution, and HTML caching — three separate capabilities in one workflow. The cached HTML persists for later analysis (useful for compliance or A/B testing). Caching incurs a small additional credit cost; confirm your plan supports it.
Use-case deep-dives
When Search API beats manual scraping for pricing intel
A 6-person growth team at a B2B SaaS company needs to track competitor pricing changes weekly across 12 rivals. They're currently screenshotting pricing pages manually. Search API's multi-engine coverage (Google, Bing, DuckDuckGo) plus HTML caching means one Switchy agent can run the same query set every Monday, store snapshots, and flag deltas without rate-limit drama. The geo-location tool matters here if competitors show regional pricing. This works cleanly up to about 30 queries per run—beyond that, the 5-tool limit starts to feel cramped because you can't batch or filter results programmatically. If your team runs this workflow more than twice a week, you'll burn through credits fast. For weekly cadence and under 50 queries total, this MCP is the right call.
Why Search API falls short for real-time support triage
A 3-person support team at a dev-tools startup wants an agent that searches their docs, GitHub issues, and Stack Overflow when a ticket comes in. Search API can hit all three sources, but the workflow breaks on latency and cost. Each support query triggers 3-5 searches (docs, then GitHub, then SO if nothing lands), and the MCP has no built-in result ranking or relevance scoring—you're parsing raw JSON every time. The cached search tools help if you're re-answering the same question, but most support tickets are unique. For this scenario, a dedicated docs search MCP or a vector store integration will close tickets faster. Search API works if you're building a weekly FAQ digest, not live triage.
When Search API wins for quarterly content planning
A 4-person content team at a fintech blog plans topics quarterly by searching what competitors and news sites are covering. They run 20-30 keyword searches across Google and YouTube, export the top 10 results per query, and tag gaps in a spreadsheet. Search API's multi-engine support and HTML caching make this a clean Switchy workflow: one agent runs the search set, another parses titles and snippets into a shared doc. The account info tool is critical here—you need to track monthly credits so you don't blow the budget mid-quarter. This breaks down if you're doing daily or weekly planning because the API cost per search adds up. For quarterly cadence and under 100 total searches, this MCP delivers the data you need without manual copy-paste.
Frequently asked
What does the Search API MCP do in Switchy?
It lets your AI agents run searches across 40+ engines — Google, Bing, YouTube, Amazon, DuckDuckGo — and retrieve structured results without leaving the chat. Agents can also check your account usage, pull cached search results by ID, and target searches to specific locations. Useful when you need real-time web data or competitive research without manual copy-paste.
Do I need to create a Search API account before connecting?
Yes. You'll need an active Search API subscription and an API key from their dashboard. Paste that key into Switchy's connection flow. The MCP checks your monthly usage and remaining credits via the Get Account Info tool, so make sure your key has valid quota before connecting.
Can the MCP search Google Images or Google Shopping?
Yes, if your Search API plan includes those engines. The Search tool supports 40+ engines, including Google Images, Google Shopping, and YouTube. Check your Search API subscription tier to confirm which engines you have access to — Switchy surfaces whatever your key allows.
Why use this MCP instead of just Googling things myself?
Speed and context. Your agent can run a dozen searches in parallel, parse the structured JSON results, and synthesise an answer — all in one chat turn. You avoid tab-switching, manual scraping, and losing context. If you only search once a week, the manual route is fine. For research-heavy workflows, the MCP saves hours.
Who on the team should connect the Search API MCP?
Whoever owns your Search API subscription and has the API key. That person connects it once in Switchy; then any workspace member can invoke the tools in their chats. Usage counts against your Search API quota, not Switchy's limits, so coordinate with your billing owner if you're on a tight credit budget.