Open Source · AI Trend & Public Opinion Radar

TrendRadar

Stop mindless scrolling — read only the news you truly care about

Aggregate trending topics/AI analysis/precise multi-channel push

Go to GitHub

Core Features

From data collection to intelligent analysis — full pipeline coverage

All-in-One Trending Aggregation

Zhihu, Douyin, bilibili, Weibo, Baidu and dozens of mainstream platforms — collect trending topics from across the web in one place

Learn more

RSS Feedsv4.5+

RSS/Atom feeds, filtered, merged and pushed together — bring RSS into the very same workflow

Learn more

Precise Content Filtering

Advanced syntax like +required, !excluded, /regex/ and @count limits gives you precise control over what gets pushed

Learn more

Trend Tracking

Track ranking changes over time, auto-detect rising and falling trends, catch turning points early

AI FilteringAI

Describe your interests in natural language; AI classifies and scores automatically, pushing only what you truly care about

Learn more

AI Analysis PushAI

Deeply analyze trending dynamics, sentiment and cross-platform connections — leave shallow information behind

Learn more

AI Multilingual TranslationAI

Translate into any language, with batch processing to save tokens — read without borders

Learn more

100+ AI Model SupportAI

Powered by LiteLLM unified interface — freely switch between DeepSeek, OpenAI, Gemini, Ollama and more

9 Push Channels

Feishu, DingTalk, WeCom, Telegram, email and more — configure once, reach every endpoint

Learn more

3 Push Modes

Daily digest / current rankings / incremental monitoring — flexibly adapt to different information habits

Learn more

Enhanced HTML Reportsv6.6+

Dark mode, search, keyboard shortcuts and wide-screen support — a thoroughly upgraded reading experience

MCP Chat AnalysisAI

17 analysis tools — query trending dynamics and cross-platform connections through natural-language conversation

Learn more

Unified Schedulingv6.0+

timeline.yaml orchestrates strategies for every time slot each day, switching automatically between weekdays and weekends

Learn more

Visual Config Editor

Edit every configuration through a graphical web interface — no more hand-writing YAML

Learn more

Flexible Multi-Target Deployment

GitHub Actions / Docker / local uv — choose whichever deployment best fits your needs

Learn more

Flexible Storage Architecture

SQLite + S3-compatible cloud storage — lightweight local or persistent cloud, your choice

Learn more

System Architecture

A three-stage data pipeline — fully automated from collection to push

NewsNow API

Real-time data from dozens of mainstream trending platforms

RSS / Atom

Custom feeds with automatic freshness filtering

01Collect
Trending data fetching RSS feed fetching Data persistence
02Analyze
Precise keyword matching AI filtering Deep AI trend analysis AI multilingual translation
03Push
HTML report generation Message dispatch scheduling Parallel push to 9 channels

Unified Scheduler

  • Unified scheduling via timeline.yaml
  • Auto-orchestrates collect / analyze / push by time slot

MCP Chat Analysis Service

  • Standalone process, read-only data queries
  • 21+ analysis tools, natural-language interaction

Data Layer

SQLiteconfig.yamltimeline.yamlfrequency_words.txtai_interests.txt

Deployment Methods

Three options covering every scenario, from zero-cost to production

GitHub Actions
  • Runs in the cloud without your own server
  • Create an independent repository from the template
  • Manage sensitive config via Secrets
  • Runs automatically on a schedule
  • Requires a check-in within each 7-day cycle
Deployment guide
Local Run (uv)
  • uv manages the Python environment automatically
  • No need to install dependencies manually
  • Supports Windows / macOS / Linux
  • Convenient for local debugging and development
  • Ideal for developers and advanced users
Deployment guide

Push Channels

Covers mainstream IM and notification services — configure once, deliver everywhere

WeCom
Personal WeChat
Feishu
DingTalk
Telegram
Email
ntfy
Bark
Slack
Generic Webhook

FAQ