<?xml version="1.0" encoding="utf-8" standalone="yes"?><rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom" xmlns:content="http://purl.org/rss/1.0/modules/content/"><channel><title>Tavily on EasyRouterAI Blog</title><link>https://blog.easyrouterai.com/tags/tavily/</link><description>Recent content in Tavily on EasyRouterAI Blog</description><generator>Hugo</generator><language>en-us</language><lastBuildDate>Tue, 08 Sep 2026 00:00:00 +0000</lastBuildDate><atom:link href="https://blog.easyrouterai.com/tags/tavily/index.xml" rel="self" type="application/rss+xml"/><item><title>How to Add Web Search to Any OpenAI-Compatible API</title><link>https://blog.easyrouterai.com/posts/openai-compatible-api-with-web-search/</link><pubDate>Tue, 08 Sep 2026 00:00:00 +0000</pubDate><guid>https://blog.easyrouterai.com/posts/openai-compatible-api-with-web-search/</guid><description>Most OpenAI-compatible gateways just route tokens. Here&amp;#39;s how to give any OpenAI-SDK app live web search by changing one line: the base_url. No tools parameter, no Tavily key, no agent loop.</description></item><item><title>Tavily Without a Second Bill</title><link>https://blog.easyrouterai.com/posts/tavily-alternatives-crypto/</link><pubDate>Tue, 08 Sep 2026 00:00:00 +0000</pubDate><guid>https://blog.easyrouterai.com/posts/tavily-alternatives-crypto/</guid><description>Tavily is the default search layer for AI agents — but it means a separate account, a separate API key, and a separate bill. Here&amp;#39;s how to get model + search in one call, on one balance you top up with cryptocurrency.</description></item><item><title>Why 'Model From China' ≠ 'Search Results From China'</title><link>https://blog.easyrouterai.com/posts/search-source-vs-model-origin/</link><pubDate>Tue, 08 Sep 2026 00:00:00 +0000</pubDate><guid>https://blog.easyrouterai.com/posts/search-source-vs-model-origin/</guid><description>DeepSeek, Qwen and GLM are open-weight and often Chinese-trained — but that doesn&amp;#39;t mean your search results have to be China-centric. Here&amp;#39;s how search source and model origin separate.</description></item></channel></rss>