Superhighway
Guides Pricing Get your free key
Practical answers to the questions we get about giving agents live web search — MCP, Python frameworks (LangChain, LangGraph, DSPy, OpenAI Agents SDK, Pydantic AI, Agno, smolagents, Google ADK, AutoGen, Semantic Kernel, CrewAI), TypeScript (Vercel AI SDK, Mastra), no-code tools, x402 pay-per-call.
Give Claude Code, Cursor, or any MCP-compatible agent live web search, news, and page reading with one command — npx -y superhighway-mcp.
How an autonomous agent discovers, pays for, and uses a web search API with no signup and no API key — the x402 pay-per-call flow, step by step.
How to give an AI agent live image search — direct image URLs, source pages, and thumbnails — using Superhighway's /images endpoint, paid per call via x402 or API key.
The /research endpoint runs a live web search and returns the top result pages as clean markdown — content an agent can use directly, for one payment of $0.005. Use domains= to restrict reading to specific trusted sources.
Two ways to give a Python AI agent live web search using Superhighway: via an MCP adapter (LangChain, LlamaIndex, CrewAI) or via a plain REST API call with an API key.
Two ways to add live web search to an OpenAI Agents SDK agent: load Superhighway as an MCP server via MCPServerStdio, or wrap the REST API as a @function_tool with an API key.
Two ways to give a Pydantic AI agent live web search: load Superhighway via MCPServerStdio (x402 wallet), or define @agent.tool_plain functions with a free API key.
Two ways to add live web search to n8n: install the native Superhighway community node (recommended) or use the HTTP Request node with an API key. Also covers Make and Zapier.
Two ways to give an Agno agent live web search: load Superhighway via MCPTools (x402 wallet, fully autonomous), or define a custom Toolkit with a free API key.
Two ways to give a Vercel AI SDK agent live web search: load Superhighway via MCP (x402 wallet, fully autonomous), or wire up typed tool() helpers with a free API key.
Two ways to give a HuggingFace smolagents agent live web search: load Superhighway via MCP (x402 wallet, fully autonomous), or define @tool functions with a free API key.
Two ways to give a Mastra agent live web search: load Superhighway via MCPConfiguration (x402 wallet, fully autonomous), or define createTool() helpers with a free API key.
Two ways to give a Google ADK agent live web search: load Superhighway via MCPToolset (x402 wallet, fully autonomous), or pass plain Python functions as tools with a free API key.
Two ways to give a Microsoft AutoGen agent live web search: load Superhighway via mcp_server_tools (x402 wallet, fully autonomous), or wrap REST calls with FunctionTool and a free API key.
Two ways to give a Microsoft Semantic Kernel agent live web search: load Superhighway via MCPStdioPlugin (x402 wallet, fully autonomous), or wrap REST calls in a @kernel_function plugin with a free API key.
Two ways to give a CrewAI agent live web search: MCPServerAdapter for all five tools (x402 autonomous payments), or @tool-decorated functions over the REST API with a free key.
Two ways to give a LangGraph agent live web search: MultiServerMCPClient (x402 autonomous payments), or @tool-decorated functions over the REST API with a free key.
Two ways to give a DSPy program live web search: wrap a function as a dspy.Tool for a ReAct agent, or build a custom dspy.Module that calls Superhighway and feeds a typed Signature.
Two paths: (1) a custom @component wrapping the Superhighway REST API with a free key; (2) MCPTool / MCPToolset for an x402 wallet — load all five tools into a Haystack Agent over npx -y superhighway-mcp.
Two no-code paths to give a Flowise agent live web search: a Custom Tool node that calls the Superhighway REST API with a free key, or an MCP node that loads all five tools (search, news, images, scrape, research) over npx -y superhighway-mcp.
Two no-code paths to give a Dify workflow or chatflow live web search: an HTTP Request node calling the Superhighway REST API with a free key, or a Custom Tool imported from Superhighway's OpenAPI spec that gives a Dify Agent node all five tools (search, news, images, scrape, research).
Connect Superhighway's web search API to any LLM via LiteLLM's unified gateway — write the tool once, run it against GPT-4, Claude, Gemini, or Llama. Function calling with a free key, or load all five tools over MCP.
Connect Superhighway's web search API to any Composio-based agent as a first-class custom tool — register it once, use it from OpenAI, LangChain, CrewAI, Claude, or any framework Composio supports.
Use Superhighway to search the live web and Instructor to extract typed, validated Pydantic models from the results — retrieval plus structured extraction in a few lines of Python.
Use Superhighway with OpenRouter to add real-time web search to any of 300+ LLMs — one setup, any model. Function calling that switches between Claude, GPT-4o, Gemini, and free models by changing a single line.
Fix empty results, 401/402 errors, rate-limit responses, slow /research calls, and other common issues with the Superhighway web search API.
Add Superhighway as the retrieval step in a RAG pipeline — no vector store required. LangChain LCEL, LlamaIndex, and standalone patterns with Python examples.
Fetch any URL as clean, LLM-ready text with a single API call. No browser needed, no JS rendering, no HTML parsing. Python and JavaScript examples with API key or x402 pay-per-call.
Give a Google Gemini app real-time web search using Superhighway and Gemini's native function calling. Python examples with google-generativeai and google-genai SDKs — zero extra infrastructure.
Give any app using the OpenAI Python SDK live web search results using Superhighway and the chat.completions tools parameter — no agent framework required.
Give any Python app using the Anthropic SDK live web search using Superhighway's tool_use integration — no agent framework required.
Production pattern for routing queries between Superhighway (live web) and a vector store — a query classifier decides which retrieval path to use based on whether the question needs current or static knowledge.
Use Mistral's function calling API with Superhighway to give your Mistral app real-time web search, news, and page scraping.
Use Groq's ultra-fast inference with Superhighway to add real-time web search, news, and page scraping to your Groq app.
A complete Python agent that fetches news on custom topics, deduplicates, and uses an LLM to write a concise daily briefing — ready to run on a schedule.
A complete Python agent that monitors websites for changes, diffs the content, and uses an LLM to summarize what changed — ready to run on a cron schedule.
A complete Python agent that uses web search, page scraping, and multi-source research to generate competitive intelligence reports automatically.
A Python agent that tracks product prices automatically: search for listings, scrape pages to clean Markdown, extract the price with an LLM, and alert when it drops below a threshold.
A Python B2B prospecting agent: take an ideal customer profile, search for matching companies, scrape their sites for contacts and team info, research deep background, then score and structure each lead into a CRM-ready CSV with an LLM.
A Python research assistant for writers, journalists, and marketers: give it a topic and it gathers recent news, finds authoritative sources, scrapes full article text, synthesizes background, then uses an LLM to produce a structured research brief with an outline, key stats, angles, and citations.
A Python agent for job seekers and recruiters: give it a search profile (role, skills, location, salary) and it finds matching job postings across the web, scrapes each full description, pulls company background, then uses an LLM to score fit, flag deal-breakers, and rank opportunities with explanations.
A Python agent for marketing teams, community managers, and PR professionals: give it a brand name and keywords, and it tracks news coverage and web mentions, scrapes the full content of key posts, then uses an LLM to score sentiment, classify urgency, and tag each mention — producing a daily digest of what's being said about you and what to do about it.
A Python agent for real estate investors, agents, and homebuyers: give it a location and property criteria, and it finds listings, scrapes full property details, researches neighborhood context (schools, safety, walkability, market trends), pulls recent local market news, then uses an LLM to generate a structured property report with an investment score and a buy/consider/pass recommendation for each home.
Python agent that researches destinations, finds attractions, pulls travel news, and generates a day-by-day itinerary using all four Superhighway endpoints.
Python agent that researches companies, scrapes financial data, pulls market news, and generates a structured investment brief using all four Superhighway endpoints.
Python agent that finds research papers, scrapes abstracts and findings, synthesizes a topic, and generates a structured literature review with research gaps and reading order.
Python agent that researches medical conditions, finds clinical trials, scrapes study details, and generates a structured health research brief with questions to ask your doctor.
Python agent that synthesizes regulatory landscapes, finds compliance requirements, scrapes regulatory pages, and generates structured compliance briefs with checklists and enforcement alerts.
Python agent that researches CVEs, threat actors, and attack techniques using live web search, then generates structured threat intelligence reports with mitigation steps and IOCs.
Python agent that researches suppliers, commodity pricing, logistics disruptions, and trade risks using live web search, generating structured supply chain briefs with risk assessments and alternative sourcing recommendations.
Python agent that researches company ESG profiles, sustainability disclosures, and regulatory requirements using live web search, generating structured briefs with E/S/G pillar breakdowns and regulatory exposure analysis.
Python agent that researches job market demand, salary ranges, key skills, and hiring trends for any role or skill using live web search, generating structured talent briefs for recruiters, HR teams, and founders.
Python agent that researches comparable M&A transactions, deal multiples, potential acquirers, and recent deal activity using live web search, generating structured M&A intelligence briefs for corporate development, PE analysts, and investment bankers.
Python agent that researches drug pipelines, clinical trial results, FDA status, and competitive landscapes using live web search, generating structured drug discovery intelligence briefs for pharma researchers, biotech analysts, and VCs.
Python agent that maps patent landscapes, finds prior art, assesses FTO risk, and tracks competitor patent filings using live web search — generating structured patent intelligence briefs for R&D teams, product managers, and IP professionals.
Python agent that researches electricity markets, renewable energy capacity, commodity prices, and energy policy using live web search — generating structured energy market briefs for analysts, developers, investors, and grid planners.
Build a Python agent that researches cryptocurrency assets and DeFi protocols using live web data — market metrics, on-chain analytics, tokenomics, and recent developments. Outputs structured JSON briefings for investment research, due diligence, and market monitoring.
Python agent that researches federal contract opportunities — agency spending, contract vehicles, NAICS codes, set-aside landscape, incumbents, and recompetes — using live web search across SAM.gov, FPDS, and USASpending for BD teams and GovCon shops.
Python agent that researches the education technology landscape — market size, key players, evidence base, FERPA/COPPA considerations, district procurement, and funding activity — using live web search across EdSurge, Crunchbase, research clearinghouses, and edtech news for investors, founders, and curriculum leaders.
Python agent that researches the media and entertainment industry — market structure, distribution landscape, content economics, recent deals, and IP/licensing — using live web search across Variety, Deadline, Hollywood Reporter, Nielsen, and Parrot Analytics for content strategists, streaming analysts, and entertainment investors.
Python agent that researches agriculture and AgTech — crop production, commodity prices, precision-ag technology, weather and climate risk, and AgTech investment — using live web search across USDA, FAO, AgFunder, and farm-news sources for commodity traders, precision-ag startups, farm managers, agricultural investors, and food supply chain analysts.
Python agent that researches teams, players, and competitions — performance trends, advanced metrics (xG, WAR, EPA, Corsi), historical context, roster and injury status, tactics, and transfer/contract situations — using live web search across Sports Reference, FBref, and recent sports news for sports analysts, team front offices, sports journalists, fantasy operators, scouting departments, and sports media.
Python agent that researches manufacturing and Industry 4.0 — production processes, OEE benchmarks, IIoT and automation technology, quality and compliance standards, workforce and skills, supply chain integration, and capex trends — using live web search across The Manufacturer, IndustryWeek, and Automation.com for plant managers, manufacturing engineers, industrial IoT teams, supply chain analysts, manufacturing investment analysts, and automation vendors.
Python agent that researches retail and e-commerce markets — category landscape, competitive positioning, channel mix (DTC vs. marketplace vs. wholesale), consumer insights, pricing and promotion, digital shelf, inventory signals, and financial dynamics — using live web search across Retail Dive, Modern Retail, and Chain Store Age for category managers, e-commerce analysts, retail buyers, DTC brand founders, retail investors, consumer goods manufacturers, and retail technology teams.
Python agent that researches insurance markets and the InsurTech ecosystem — market structure, carrier landscape, underwriting and loss trends, regulatory environment, technology adoption, distribution, and funding — using live web search across Insurance Journal, Carrier Management, and PropertyCasualty360 for insurance carriers and underwriters, InsurTech founders and investors, reinsurance analysts, actuarial teams, brokers and MGAs, corporate risk managers, and insurance compliance teams.
Python agent that researches automotive markets and the EV transition — OEM landscape, EV adoption metrics, battery and charging technology, regulatory and policy environment, supply chain, dealer dynamics, and financial signals — using live web search across WardsAuto, Automotive News, Motor Trend, and Car and Driver for automotive industry analysts, EV startup founders and investors, OEM strategy teams, automotive engineers and product managers, fleet operators, charging infrastructure investors, and automotive policy researchers.
Python agent that researches HR technology markets and people analytics — HRIS/HCM platforms, compensation intelligence, performance and learning tools, DEI analytics, and workforce trends — using live web search across HR Executive, HR Technologist, SHRM, and Gartner for CHROs and VP People, HR tech vendors and their buyers, people ops teams, HR technology investors, workforce management consultants, compensation analysts, and HR software implementation consultants.
Python agent that researches the PropTech and real estate technology ecosystem — property management and leasing platforms, CRE brokerage tech, smart buildings/IoT, construction technology, and real estate fintech — using live web search across Propmodo, The Real Deal, Bisnow, and NAIOP for PropTech investors, CRE tech buyers, real estate developers, property management companies, smart building/IoT teams, construction technology companies, and real estate fintech analysts.
Python agent that researches biotech pipelines and the clinical trial landscape — Phase I/II/III trial data, FDA regulatory pathways, competitive pipeline intelligence, and biopharma M&A/licensing — using live web search across ClinicalTrials.gov, STAT News, Fierce Pharma, and Endpoints News for biotech/pharma investors, competitive intelligence teams, clinical researchers, biotech founders, BD teams, healthcare analysts, and life sciences consultants.
Python agent that researches the semiconductor and chip industry — fab capacity, process nodes, competitive landscape, CHIPS Act and export controls, and investment signals — using live web search across EE Times, SemiEngineering, and IEEE Spectrum for chip investors, fabless companies, EDA vendors, semiconductor analysts, and geopolitical researchers.