GEOInflow Open-Sourced for AI Crawler and GEO Analytics

GEOInflow, a self-hosted web and AI crawler analytics system for Generative Engine Optimization, was released under the MIT license with one-click Docker deployment. It tracks AI search referrals, crawl logs, content coverage and conversions, and identifies 61 AI crawler identities across platforms including OpenAI, Anthropic, Perplexity, Google, ByteDance, Baidu, Kimi and Tencent Hunyuan. The project is jointly published by Liekan Technology and Beijing Daily Zhiyuan GEO, built on Umami, and initiated by Yao Jingang, Xiangyang Qiaomu and Ren Qiang.
Event version 1.
Key insights
AI crawler identification
GEOInflow identifies which AI platforms crawl a site, including OpenAI, Anthropic, Perplexity, Google, ByteDance, Baidu, Kimi and Tencent Hunyuan, and distinguishes training, AI search and user-triggered visits. It shows which pages were crawled and which key pages have never been crawled.
GEO performance tracking
The system tracks sessions, landing pages and conversions referred by AI search, allowing teams to measure the actual effect of GEO. It also reports crawl errors and content coverage gaps.
Full web analytics included
GEOInflow provides complete website statistics including PV/UV, event funnels, heatmaps, retention, team permissions and sharing. Core statistics require no model API key, use cookie-free collection by default, and keep data on the user's own server.
Crawler taxonomy and dashboards
The system organizes 61 different AI crawler identities and classifications worldwide and builds multiple new data analysis dashboards on that basis.
Open-source basis and publishers
GEOInflow is developed based on Umami and adopts the MIT open-source license. It is jointly published by Liekan Technology and Beijing Daily Zhiyuan GEO, with project initiators Yao Jingang, Xiangyang Qiaomu and Ren Qiang.