Keenable
Keenable is a research-driven AI infrastructure company building independent web-search and knowledge-access systems for large-scale AI applications. Its goal is to make web-scale information accessible to models and agents through low-latency retrieval primitives rather than human-oriented ‘ten blue links’ search.
At a Glance
- AI laboratories
- Inference platforms and model builders
- AI-agent developers
- Researchers
- +2 more
AI Tools by Keenable
(1)
Keenable
Web Search API for AI Agents
Discussions
No discussions yet
Be the first to start a discussion about Keenable
Latest News
cognee’s agent memory now reads the web through Keenable
Keenable partners with Baseten to make web search native to inference
Keenable introduces SELECT, a SQL-like interface for querying the web as a database
Keenable now runs inside Lightpanda
Products & Services
Agent-oriented web search and page-content access over Keenable’s independent 100B+ document index. The API exposes search_web_pages and fetch_page_content, and is available through REST, MCP and the CLI.
A SQL-like, agentic web-search interface that queries broad slices of the web, extracts structured columns, and supports filtering, grouping, aggregation and cited typed rows.
A point-in-time search API for searching prior versions of web pages and retrieving historical snapshots; query_time rewinds both the corpus and ranking.
A live, open-source benchmark for measuring agentic web-search quality across news, finance, scholarly, rare-entity and legal queries.
Market Position
Keenable positions itself as independent, AI-native web-search infrastructure rather than a human-oriented search API or enterprise-search product. It emphasizes web-scale coverage, low latency, cost efficiency, dedicated/on-premises capacity and retrieval interfaces such as SELECT; its disclosed comparison set includes Google/Serper, Bing/SearchAPI, Brave, Tavily, Parallel, Exa and Perplexity Search as Code.
Leadership
Founders
Andrey Styskin
Co-founder and CEO; previously led Yandex’s Search and Ads business and its search, AI and cloud division, and later served as Director of Web Infrastructure at Amazon AGI.
Matthias Petri
Co-founder and Chief Scientist; previously a Principal Applied Scientist at Amazon AGI, where he built the web-grounding service behind Alexa.
Executive Team
Andrey Styskin
Co-founder & CEO
Former CEO/leader of Yandex Search and Ads and former Director of Web Infrastructure at Amazon AGI.
Matthias Petri
Co-founder & Chief Scientist
Former Principal Applied Scientist at Amazon AGI; built the web-grounding service behind Alexa.
Founding Story
Keenable was started after Andrey Styskin and Matthias Petri saw at Amazon that AI applications needed web access designed for agents rather than human search. They set out to build an independent, web-scale index and new retrieval primitives so agents could query current and niche knowledge cheaply, quickly and repeatedly.
Business Model
Revenue Model
Usage-based API pricing for web-search and page-fetch requests, with dedicated-capacity and enterprise cloud/on-premises offerings for AI labs and inference platforms.
Pricing Tiers
Pay-as-you-go, cloud-only access for agent builders and enthusiasts; includes 100,000 free requests per month.
Dedicated capacity for AI-scale customers, with cloud and on-premises access.
Target Markets
- AI laboratories
- Inference platforms and model builders
- AI-agent developers
- Researchers
- Voice-AI developers
- Companies needing web-scale retrieval during model training or inference
- Grounding LLM and agent responses with current web information
- Web search during model training and runtime inference
- AI-agent research and deep retrieval
- Voice agents needing low-latency live retrieval
- Structured market, pricing and competitive research over many web pages
- Agent memory and knowledge-graph ingestion