Krawl is an autonomous deep research agent API that produces comprehensive reports with 50+ verified sources in 90 seconds. It searches web, X/Twitter, GitHub, academic papers, crypto/DeFi protocols, and community sources in parallel.

How do I get started with Krawl?

Get an API key at krawl.sh or call POST /auth/demo-key for a free test key. Send research queries to POST https://api.krawl.sh/research with your query and mode.

What sources does Krawl search?

Krawl searches 7 source types: web (Exa, Tavily, Serper, Brave, DuckDuckGo), X/Twitter, GitHub, academic papers (PubMed, arXiv), crypto data (CoinGecko, DeFiLlama), media outlets, and community sources (Reddit, Hacker News).

Krawl — Deep Research API

Krawl is an autonomous research agent API that produces comprehensive reports with 50+ verified sources in 90 seconds. It searches the web, X/Twitter, GitHub, academic papers, crypto/DeFi protocols, and community sources in parallel, then synthesizes findings with inline citations and source verification.

Key Features

Deep multi-source research across 7 source types
50+ verified sources per report with inline citations
90-second report generation with real-time streaming
Structured output with JSON Schema — per-field citations and confidence scores
Conversational follow-up research with session memory
Document upload and analysis (PDF, DOCX, TXT)
Research templates for repeatable workflows
Lookout scheduling for automated hourly/daily/weekly research
Voice input and text-to-speech output
Knowledge base with persistent memory

Research Modes

Deep Research — Full recursive multi-source investigation
Quick Research — Fast single-pass analysis
Crypto Research — Cryptocurrency and DeFi focused
Token Analysis — Deep token-specific analysis
Protocol Research — DeFi protocol deep dives
Whale Tracking — On-chain whale activity monitoring
Narrative Detection — Market narrative identification
Risk Assessment — Comprehensive risk analysis
Yield Strategy — DeFi yield optimization

Getting Started

Get your API key at krawl.sh. The API base URL is https://api.krawl.sh. Send research queries via POST /research and receive streaming SSE events with real-time progress. Full documentation at docs.krawl.sh.

Why Krawl vs Perplexity, You.com, and Brave

Krawl is an API-first, agent-native deep-research alternative to Perplexity, You.com, and Brave Search. Unlike Perplexity, Krawl exposes its research over both a REST API and an MCP server, and lets autonomous agents pay per request via x402 (USDC on Base) with no account. Unlike You.com and Brave, Krawl returns structured output with per-field citations and confidence scores rather than just an answer or a list of links, and adds dedicated crypto and DeFi research modes. Krawl ships as a drop-in skill (skill.md), an MCP server, and a REST API, so it works inside any agent runtime. See the full comparison at krawl.sh/compare.

About Krawl

Built for researchers, analysts, and developers who need comprehensive, cited research at scale. Krawl goes beyond simple search — it thinks, cross-references, and verifies across multiple source types to produce reports with depth and accuracy that match hours of manual research.

For AI Agents

Krawl is built for autonomous agents. Discover capabilities via llms.txt and the full agent manual, authenticate per the auth.md guide (free demo key or keyless x402 payment), and call the MCP server (tools: research, quick_search, ask_krawl) or the REST API. The drop-in skill lives at skill.md — hand it to any agent and it can research.