← CookbookView source on GitHub ↗

Tavily + Token Factory

Tavily

About Tavily

Tavily is an LLM-optimized search API built specifically for agents. It delivers clean, structured, and context-aware web results — making it ideal for research, competitive intelligence, and multi-step agentic workflows. Tavily handles query formulation, result filtering, and content extraction so your Nebius-powered agents can focus on reasoning and synthesis instead of wrestling with raw HTML. Many examples in this cookbook pair Tavily search with Nebius Token Factory models to build agents that plan, research, and produce decision-grade outputs.

Here are some examples of Tavily and Token Factory.

Example Description Tech Stack
Tavily tool calling Jupyter notebook demonstrating Tavily tool-calling with Nebius Token Factory Token Factory · Tavily
Tavily search Agent Search with Tavily Token Factory · Tavily
competitive research agent Deep agent example that uses 3 sub-agents to do pricing-research, news-research and sentiment-research LangChain deepagents · Nebius TF · Tavily
Universal Deep research Multi-step research pipeline with FastAPI backend + Next.js frontend; fetches web results via Tavily FastAPI · Next.js · Nebius TF · Tavily