TrendRadar is an open-source, AI-driven trend and public-opinion monitor with more than 63,000 stars on GitHub that aggregates hot lists from many platforms and pushes AI-filtered briefings to your phone.
Information overload is the problem TrendRadar attacks: instead of doom-scrolling a dozen trending pages, it continuously collects hot topics from multiple platforms plus your own RSS feeds, filters them with precise keyword rules, and then uses an LLM to screen stories, translate, analyze sentiment and trends, and compile a briefing. Alerts and reports are pushed through the channels people actually read — WeChat, Feishu, DingTalk, Telegram, email, Slack, ntfy and Bark. It also exposes an MCP server, so AI assistants can query the collected data conversationally for analysis and forecasting. Everything is self-hostable with Docker and your data stays on your own infrastructure. The project has trended 51 times, is GPL-3.0 licensed, and was actively developed in October 2026.
TrendRadar is particularly strong in the Chinese information ecosystem, where the stories that matter often start on domestic hot lists rather than global feeds. By combining structured aggregation with LLM judgment — deduplicating, translating and ranking by relevance instead of raw heat — it turns the daily chaos of trending pages into a briefing a person can finish in five minutes.
TrendRadar is free and open source under the GPL-3.0 license. You provide your own LLM API access for the AI analysis features; hosting can run on a small home server.
Media watchers, marketers, researchers and news junkies — especially in the Chinese platform ecosystem — who want signal instead of noise from the daily flood of trending lists.