From Nebius Inflection to new ways to access Tavily, here’s what’s new this month.
By Tavily TeamJul 2, 2026

Calling a tool means sending your data to someone else's servers and what happens there is what nobody checks. Keyless means no terms, no leverage, at exactly the moment you've done the least to govern your data. The fix isn't avoiding keyless. It's knowing when to graduate off it.

How a product leader rebuilt AI research on Tavily, turning a 5-hour manual cycle into a 25-minute agent that pulls live, cited web sources and verifies every citation before it's seen.

Snowflake Summit, new integrations (x402) and agent deployments across the ecosystem.

Tavily Search is now live on x402, letting AI agents pay for fresh web search at runtime using only a USDC wallet on Base. No API keys, signup flows, or preconfigured accounts required.

At LangChain Interrupt, we demoed an agent that uses Tavily for real-time search and Nebius for scalable inference to continuously gather, verify, and structure competitive intelligence.

Peerbound needed accurate data on hundreds of thousands of companies. After trying Perplexity and Exa, they switched to Tavily's AI Search API — and went from 2/10 to 10/10 accurate results.

Hermes Agent is a self-improving, persistent AI runtime and its web backend influences how it learns. Here’s how Tavily plugs in to power higher-quality search, extraction, and research inside the loop.

April Updates: /research is becoming a building block for modern AI systems, alongside new agent skills, real-world benchmarks, and deeper integrations across the ecosystem.

Real-time web intelligence meets enterprise-grade agent execution on Arcade.dev. Build agents that don’t just think, but securely act across your tools and workflows.