Find and compare
Benchmarks and prices
Releases and labs
Learn and sources
Model Directory
Tracked models by provider, price, and status
Leaderboard
Frontier, value, speed, and evaluated views
Release Desk
Newly tracked launches and briefs
Benchmark Library
External eval sources and caveats
Methodology
How AIRH scores and evidence states work
Reliability Floor
Everyday trust basics
News Desk
Latest captured AI signal
Updates
Data freshness and site changes
Changelog
Recent hub activity
LLM Compare
Prices, speed, context, and quality
Head-to-Head
Compare two models directly
Best Value
Quality per blended cost
Guides
Practical explainers and buying decisions
AI Academy
Structured learning path
Prompt Library
Reusable prompt patterns
Machine-assisted translation
这个新闻中文层目前是机器辅助阅读层,保持 `noindex`。每日动态和数据变化仍以英文新闻页为主。
这条路由的意义不是“假装已经全部翻译完成”,而是把动态新闻的正确处理方式先落地:可读、诚实、可持续。