The web, Indexing for AI Agents

Powering the future of AI with fast, high-quality and structured web search.

WhatisthecurrentpriceofBitcoin?

Index: Web-01

Thousands of companies search the web with Serpex

NotionCursorAWSDatabricksGroqGammaMonday.comHubSpotLovablePoint72NotionCursorAWSDatabricksGroqGammaMonday.comHubSpotLovablePoint72NotionCursorAWSDatabricksGroqGammaMonday.comHubSpotLovablePoint72

Frontier search powered by our AI research lab

Serpex's search engine is developed in-house, allowing for full controllability.

Independent Search Engines

[Low Latency <1s]

Serpex Fast

Perplexity

Brave

Methodology

We use a SingleStep harness that searches prompts directly with default settings. We pass the query and 10 search results to GPT-5 mini, with GPT-4o-mini grading the output for correctness.

FRAMES ↗

A dataset released by Google designed to test Retrieval-Augmented Generation (RAG) applications on factuality, retrieval accuracy, and reasoning.

Tip-of-the-Tongue

A benchmark for evaluating search engines on queries where users remember partial information about what they're looking for.

Seal0 ↗

Evaluates web search systems specifically against conflicting, noisy, and ambiguous information.

SimpleQA ↗

A factuality benchmark released by OpenAI that measures the ability for language models to answer short, fact-seeking questions.

Infrastructure

Engineered for unmatched scale.

0+

Web pages indexed by Serpex (and counting)

State-of-the-art
Accuracy

Flexible
Pricing

Sub-second
Latency

99.9%
uptime, SLA-
backed

End-to-End Encryption

Single Sign On

Zero Data Retention

Enterprise
Security

Developer API

Search at the Speed of Thought

Request

curl 'https://api.serpex.io/v1/search' \
  -H 'Authorization: Bearer sk-...' \
  -G \
  --data-urlencode 'ns=knowledge.medical' \
  --data-urlencode 'q=Phase III trials GLP-1 agonists active enrollment' \
  --data-urlencode 'fields=nct_id,trial_phase,condition_name,issuing_org' \
  --data-urlencode 'limit=5'

Response

{
  "namespace": "knowledge.medical",
  "total_hits": 847,← namespace-aware routing
  "latency_ms": 71,
  "results": [
    {
      "nct_id": "NCT05536141",
      "trial_phase": "Phase III",← typed schema fields — not "snippet"
      "condition_name": "Type 2 Diabetes Mellitus",
      "drug_name": "semaglutide",
      "drug_class": "GLP-1 receptor agonist",
      "trial_status": "Recruiting",
      "issuing_org": "WHO",
      "evidence_level": "Grade A",← domain authority score built in
      "mesh_terms": ["Semaglutide", "Glucagon-Like Peptide-1 Receptor"],← taxonomy-specific metadata
      "relevance_score": 0.94
    }
  ]
}

Our Platform

Build the future of AI apps

with Serpex

We are developing the right tools for your team across the entire AI lifecycle.

Retrieval-Augmented Agents

Power your RAG pipelines with high-recall web indexing and structured entity extraction.

Voice-Powered Search

Enable natural language search queries for voice agents with low-latency retrieval and indexing.

Semantic Search APIs

Go beyond keyword matching with vector-based search that understands user intent and context.

Context-Aware Retrieval

Deliver hyper-relevant search results by indexing the web with deep semantic understanding.

Market Intelligence Search

Search through millions of SEC filings, earnings calls, and news articles with structured data output.

Medical Literature Retrieval

Index and search clinical trials, research papers, and patient data with high precision and security.

Case Law Discovery

Perform deep legal research across jurisdictions with intelligent indexing of statutes and case law.

Codebase Semantic Search

Index massive repositories and documentation to provide intelligent code search for developer tools.

Web Index

Industry-leading web index built for agents

ENTITIES

entities.*

"What IS something?"

entities.people ──┘←structured person records

entities.organizations ──┘←company + institution records

entities.products ──┘←product specs + metadata

entities.places ──┘←geo + venue records

entities.events ──┘←time-bounded occurrences

KNOWLEDGE

knowledge.*

"Explain something in depth"

knowledge.web ──┘←long-form web documents

knowledge.research ──┘←academic papers + preprints

knowledge.legal ──┘←case law, statutes, patents

knowledge.medical ──┘←clinical guidelines, drug data

knowledge.government ──┘←policy, budgets, filings

US-East Index

US-West Index

EU-Central Index

AP-East Index

AP-South Index

SA-East Index

AP-Southeast Index

EU-North Index

EU-West Index

IN-West Index

420k pages/s

380k pages/s

290k pages/s

185k pages/s

156k pages/s

134k pages/s

STREAMS

streams.*

"What is happening RIGHT NOW?"

└── streams.news→articles, press releases

└── streams.finance→earnings, analyst reports

└── streams.social→Reddit, HN, StackOverflow

CODE

code.*

"Show me how to build something"

└── code.repos→GitHub/GitLab repositories

└── code.packages→npm, PyPI, crates.io

└── code.docs→API references, guides

MEDIA

media.*

"Show me something visual or auditory"

└── media.video→video metadata + transcripts

└── media.audio→podcasts, speeches

└── media.images→images with rich context

Engineering Excellence

System Architecture

Billion-Scale Engine

A custom distributed indexing architecture capable of ingesting and querying billions of documents in real-time.

Read Research

Content Processing

Our proprietary pipeline employs advanced heuristics to filter noise and ensure high-quality indexing.

Read Research

Sub-Millisecond Latency

Relentless networking optimizations achieve sub-millisecond query routing and unprecedented response times.

Read Research

Loved by AI-native companies

See why the top AI companies are switching to Serpex for their retrieval needs.

"Serpex completely changed how we train our models. The web index is incredibly comprehensive and the data is always fresh."

Alex Chen, CEO at TechCorp

"The search latency is mind-blowing. We needed real-time retrieval for our agents, and Serpex delivered beyond expectations."

Dan Miller, CTO at SecureNet

"We replaced our entire crawling infrastructure with Serpex. It saved us hundreds of engineering hours and significantly reduced costs."

Stephanie Lin, COO at InnovateCo

"The structured entities make it so easy to extract company and product data. Our AI agents are now 10x more accurate."

Marie Dubois, Head of AI at FuturePlanning

"Hands down the best search API on the market. The documentation is crystal clear and integration took literally minutes."

Andre Gomez, Lead Engineer at CreativeSolutions

"Serpex handles the scale of our queries effortlessly. We process millions of searches daily without a single hiccup."

Jeremy Scott, Product Manager at TimeWise

"The relevance of the search results is unmatched. Our RAG pipelines have never been more accurate and reliable."

Pam Beesly, AI Researcher at BrandBuilders

"Access to real-time news streams has completely transformed our financial analysis tools. It's a game changer for us."

Daniel Kim, Data Scientist at AnalyticsPro

"Finally, a search engine built specifically for developers and AI agents. It's exactly what the industry has been waiting for."

Fernando Ruiz, Founder at UserFirst

"The quality of the code documentation index is fantastic. It powers our coding assistant and the results are always spot on."

Andy Bernard, DevOps Engineer at CloudMasters

Frequently Asked Questions

We've Got the Answers You're Looking For

Have any more questions? Contact us

A traditional search engine treats its index as one giant flat table of documents, which works for humans scanning results. AI agents, however, need structured data. The data lake model classifies the web into zones (Entities, Knowledge, Streams, Code, Media) so AI systems get exactly the typed, structured knowledge they need.

Start building today

Build something that

knows

things.

1,000 free queries to get started. No credit card required.

Frontier search powered by our AI research labFrontier search powered by our AI research lab

Independent Search Engines

Methodology

FRAMES ↗

Tip-of-the-Tongue

Seal0 ↗

SimpleQA ↗

0+

State-of-the-artAccuracy

FlexiblePricing

Sub-secondLatency

99.9%uptime, SLA-backed

EnterpriseSecurity

Retrieval-Augmented Agents

Voice-Powered Search

Semantic Search APIs

Context-Aware Retrieval

Market Intelligence Search

Medical Literature Retrieval

Case Law Discovery

Codebase Semantic Search

ENTITIES

KNOWLEDGE

STREAMS

CODE

MEDIA

Billion-Scale Engine

Content Processing

Sub-Millisecond Latency

"Serpex completely changed how we train our models. The web index is incredibly comprehensive and the data is always fresh."

"The search latency is mind-blowing. We needed real-time retrieval for our agents, and Serpex delivered beyond expectations."

"We replaced our entire crawling infrastructure with Serpex. It saved us hundreds of engineering hours and significantly reduced costs."

"The structured entities make it so easy to extract company and product data. Our AI agents are now 10x more accurate."

"Hands down the best search API on the market. The documentation is crystal clear and integration took literally minutes."

"Serpex handles the scale of our queries effortlessly. We process millions of searches daily without a single hiccup."

"The relevance of the search results is unmatched. Our RAG pipelines have never been more accurate and reliable."

"Access to real-time news streams has completely transformed our financial analysis tools. It's a game changer for us."

"Finally, a search engine built specifically for developers and AI agents. It's exactly what the industry has been waiting for."

"The quality of the code documentation index is fantastic. It powers our coding assistant and the results are always spot on."

We've Got the Answers You're Looking ForWe've Got the Answers You're Looking For

Frontier search powered by our AI research lab

State-of-the-art
Accuracy

Flexible
Pricing

Sub-second
Latency

99.9%
uptime, SLA-
backed

Enterprise
Security

We've Got the Answers You're Looking For