Sailor is a search API designed for AI applications. Instead of returning raw HTML or scraped text, it gives you clean markdown, structured fields, citations, and platform-aware results — in one call.

How is Sailor different from Serper, SerpAPI, or Brave?

Those tools wrap a single SERP engine and hand you raw JSON. Sailor sits a layer above: it routes across engines, scrapes pages into clean markdown, hits social platforms natively, and returns token-efficient context built for LLMs.

How does Sailor compare to Exa or Perplexity?

Exa is great at neural recall, Perplexity at synthesis. Sailor focuses on context delivery — clean, structured, citable, model-agnostic. Bring your own LLM; Sailor handles the messy retrieval underneath.

Can I use Sailor as a drop-in replacement for Serper?

Yes. Our /v1/search endpoint accepts a Serper-compatible payload and returns a superset of its response. Most teams migrate in under 30 minutes.

What does 'token-compact' actually mean?

Sailor strips boilerplate, dedupes near-identical paragraphs, and converts content to compressed markdown. On typical pages we see 60–75% fewer tokens than raw HTML — directly cutting your model spend.

Which social platforms are supported?

X (Twitter), Reddit, YouTube, TikTok, Hacker News, and LinkedIn at launch. Each has its own ranking signals and time filters. Mastodon and Bluesky are next.

How do you handle rate limits and credits?

New accounts start with 3,000 launch credits: a 2,500-credit one-time bonus plus 500 monthly credits. Paid plans use predictable credit packages, and enhanced operations like Markdown fetch, compression, enrichment, and deep search are quoted before expensive work starts.

Do you train on my queries?

No. We don't train models on customer queries or results. Pro and Scale plans include zero-retention mode for full compliance.

AI-native search API

Search built for
agents, not browsers.

Sailor turns the open web into clean markdown, structured data, and platform-aware results — in one API call. Compact tokens, deep enrichment, zero markup mess.

Read the docs

Drop-in replacement for Serper · OpenAI tool-compatible · 99.99% uptime

sailor - search demo

live response

request.curl

POST /v1/search

$ curl https://api.sailorsearch.dev/v1/search \
  -H 'Authorization: Bearer sk_live_***' \
  -H 'Content-Type: application/json' \
  -d '{
    "q": "langchain agents",
    "format": "markdown",
    "max_tokens": 1200
  '}

One call. Search, fetch, clean, enrich.

response

200 OK

284 ms - 412 tokens

{
  "query": "langchain agents",
  "engine": "sail.blend",
  "results": [
    {
      "title": "Agents | LangChain",
      "url": "https://python.langchain.com/docs/agents",
      "markdown": "# Agents\nAgents use an LLM to choose tools...",
      "tokens": 412,
      "crawled_at": "2026-04-21T08:14:00Z"
    }
  ]
  "usage": { "tokens": 412, "latency_ms": 284 }
}

Trusted by builders shipping real AI

OpenAIVercelLangChainAnthropicCloudflareSupabaseZapiern8nHugging FaceOllama

AtSailor,we'rerebuildinghowAI🧭readstheopenweb.

Westripthemarkup,thetrackers,andthenoisesoyourmodelonlyseeswhatmatters✨

OneAPI.Everyengine,everyplatform,everypage🌊deliveredasifaseniorengineerwrotetheparserforyou.

The all-in-one search layer for AI

Stop stitching together SERP scrapers, page parsers, and social fetchers. Sailor is one API for everything your model needs to read.

Clean markdown scraping

No more <div> soup. Sailor returns LLM-ready markdown with headings, lists, code blocks and tables preserved — every time.

Structured deep search

One query fans out across engines, follows links, and returns enriched, citation-backed JSON.

Sail Engine routing

Sail Engine blends the right sources behind one stable API, with advanced routing controls available on paid plans.

Social-native search

Native endpoints for X, Reddit, YouTube, TikTok, Hacker News and LinkedIn. Real signals, real time.

Tool-ready schemas

Typed payloads, predictable errors, and agent-friendly response shapes for production workflows.

Citations & timestamps

Every chunk carries its source URL and crawl time. Build trustworthy, auditable agents.

Token-compact output

~70% smaller than raw HTML. Lower context cost, higher recall, faster inference.

How Sailor works

Three focused steps. One endpoint. Clean web context for agents without the scraping glue.

Send a query

Drop in a URL, topic, or research prompt. Use Sail Engine by default, then choose the sources and output format your agent needs.

"langchain agents" format: markdown

Sailor searches

Sailor fans out across search engines, follows the useful pages, dedupes results, and extracts the signal.

Use clean output

Get compact markdown or structured JSON with citations, timestamps, token counts, and source provenance.

200 OK

"markdown": "# Agents..."

"tokens": 412

"sources": [3]

"crawled_at": "2026-04-21"

JSON + Markdown

Ready for any model

One endpoint. Every shape you need.

Test queries from your terminal, editor, or notebook. Same payload, same response.

sailor-sdk

SDK quickstart

from sailor import Sailor

sailor = Sailor(api_key=os.environ["SAILOR_KEY"])

result = sailor.search(
    q="best vector db for rag 2026",
    engine="sail",
    social=["reddit", "hn"],
    format="markdown",
    max_tokens=4000,
)

response.json

412 ms

{
  "query": "best vector db for rag 2026",
  "results": [
    {
      "title": "Vector Databases for RAG",
      "markdown": "# Comparison\\nPinecone, Qdrant...",
      "tokens": 3214
    }
  ],
  "engine": "sail.blend"
}

8 results3,214 tokens412 msSail Engine - reddit - hn

The obvious choice for AI teams

A quick look at how Sailor stacks up against the search APIs developers reach for today.

	Sailor	SSerper	exaExa	APISerpAPI	BBrave	PPerplexity
Clean markdown output
Structured deep search
Multi-engine routing
Social-native search
Token-compact responses

Sailor

Clean markdown output
Structured deep search
Multi-engine routing
Social-native search
Token-compact responses

Serper

Clean markdown output
Structured deep search
Multi-engine routing
Social-native search
Token-compact responses

Exa

Clean markdown output
Structured deep search
Multi-engine routing
Social-native search
Token-compact responses

SerpAPI

Clean markdown output
Structured deep search
Multi-engine routing
Social-native search
Token-compact responses

Brave Search API

Clean markdown output
Structured deep search
Multi-engine routing
Social-native search
Token-compact responses

Perplexity API

Clean markdown output
Structured deep search
Multi-engine routing
Social-native search
Token-compact responses

See full comparisons

RAG pipelines

AI agents

SaaS

Research tools

Browser agents

Works for any AI workload

RAG pipelines

Agents

Research

AI agents

Lead generation

SaaS

Research tools

Brand monitoring

Browser agents

SEO platforms

Built on

Our search DNA

Four principles we refuse to compromise on. They're why your agent will never get back HTML soup again.

Clean by design

Markdown, JSON, or plain text. Never markup soup. Never surprises.

Built for agents

Tool schemas, streaming, budget caps. The primitives agents actually need.

Source-aware

Citations, timestamps, and provenance baked into every response.

Privacy-first

We don't train on your queries. Zero retention plans available on Pro.

Works seamlessly with your stack

Native SDKs and one-line tool definitions for the frameworks you already ship with.

Simple, predictable pricing

3,000 launch credits to start, then monthly packages aligned with AI search workloads.

Free

$0forever

3,000 launch credits for first-time signup.

2,500 one-time bonus credits
500 monthly credits
Sail Engine only
Community support

Request access

Starter

$29/ month

For first production agents and steady prototypes.

7,000 monthly credits
$4.14 per 1k basic searches
Paid advanced routing controls
Markdown and compression operations

Request Starter

Pro

POPULAR

$79/ month

For teams shipping real AI products.

22,000 monthly credits
$3.59 per 1k basic searches
Advanced routing and deep search
Priority support

Request Pro

Scale

$249/ month

Higher-volume AI search with room for bursts.

85,000 monthly credits
$2.93 per 1k basic searches
Higher QPS and priority routing
Volume top-ups available

Request Scale

Questions, answered

Still curious? Email us at inquiry@sailorsearch.dev — we usually reply within hours.

Give your agent the search it deserves.

Start with clean markdown, citations, social context, and structured search in one API call.

3,000 launch creditsNo credit cardLive in under 5 minutes

By signing up you agree to our terms. We'll never train on your queries.

Search built foragents, not browsers.

The all-in-one search layer for AI

Clean markdown scraping

Structured deep search

Sail Engine routing

Social-native search

Tool-ready schemas

Citations & timestamps

Token-compact output

How Sailor works

Send a query

Sailor searches

Use clean output

One endpoint. Every shape you need.

The obvious choice for AI teams

Works for any AI workload

Works for any AI workload

Our search DNA

Clean by design

Built for agents

Source-aware

Privacy-first

Works seamlessly with your stack

Simple, predictable pricing

Questions, answered

What is Sailor?

How is Sailor different from Serper, SerpAPI, or Brave?

How does Sailor compare to Exa or Perplexity?

Can I use Sailor as a drop-in replacement for Serper?

What does 'token-compact' actually mean?

Which social platforms are supported?

How do you handle rate limits and credits?

Do you train on my queries?

Give your agent the search it deserves.

Search built for
agents, not browsers.