Tavily

See also: Retrieval-augmented generation, AI agent, and Information retrieval

Introduction

Tavily is a search engine and API platform built specifically for AI agents and retrieval-augmented generation (RAG) systems. Unlike traditional search engines designed for human users, Tavily optimizes every part of its pipeline for consumption by large language models (LLMs), returning structured, ranked, and relevance-scored results that can be fed directly into AI workflows. The platform provides a suite of APIs, including Search, Extract, Map, Crawl, and Research endpoints, that let developers connect their AI applications to real-time web data through a single integration ^[1][2]. As of mid-2026, Tavily says it is trusted by more than 2 million developers ^[24].

Tavily was founded in late 2023 by Rotem Weiss and Assaf Elovic and is headquartered in New York City, with additional offices in Tel Aviv and Abu Dhabi ^[3][21]. The company grew out of an open-source project called GPT Researcher, which Elovic created in 2023 to fetch real-time web data for LLMs before ChatGPT had internet connectivity ^[3]. By August 2025, Tavily had raised $25 million in total funding, including a $20 million Series A led by Insight Partners and Alpha Wave Global ^[3]. In February 2026, Nebius, an AI infrastructure company, announced an agreement to acquire Tavily for an initial $275 million, with the deal potentially rising to $400 million based on performance milestones ^[4][5].

Tavily serves Fortune 500 enterprises and leading AI companies, including IBM, Amazon Web Services, MongoDB, Cohere, Groq, JetBrains, Monday.com, and LangChain ^[5][6]. It integrates natively with popular AI frameworks such as LangChain, LlamaIndex, and supports the Model Context Protocol (MCP) for direct connectivity with AI assistants ^[7][8].

How does Tavily differ from a traditional search engine?

Traditional search engines and their APIs are built to return a ranked list of links for a human to click and read. Tavily is built for machines: it aggregates content from multiple sources per query, uses proprietary AI models to score and filter that content for relevance, and returns clean, structured text that a language model can consume directly without additional scraping or parsing ^[1][2]. The stated goal is to remove the retrieval and cleaning work that developers would otherwise build themselves. "We're on a mission to onboard the next billion AI agents to the web," said Rotem Weiss, Tavily's co-founder and CEO ^[21][22].

Three properties distinguish Tavily from a conventional SERP (search engine results page) API:

• LLM-ready output. Results are returned as ranked, relevance-scored JSON snippets with optional Markdown-formatted raw content, rather than raw HTML or bare link lists ^[1][11].
• Single-call retrieval. One request can search, rank, extract, and summarize across up to 20 sources, collapsing a multi-step pipeline into a single API call ^[1][2].
• A safety layer. Tavily positions itself as "a firewall between your agents and the internet," filtering prompt-injection attempts and other hostile content before it reaches the model ^[2][15].

This machine-first design is why Tavily is frequently adopted as the retrieval backend for RAG pipelines and autonomous AI agents rather than as a consumer search product.

History

Origins: GPT Researcher

Tavily traces its roots to an open-source project called GPT Researcher, created by Assaf Elovic in 2023. At the time, ChatGPT and other consumer-facing LLMs lacked the ability to access live web data. Elovic, a software engineer who had previously served as Head of R&D at Wix and founded Tiv.ai, built GPT Researcher as an autonomous agent that could conduct deep research on any topic by searching the web, aggregating information from multiple sources, and generating comprehensive reports ^[9][3]. Weiss later recalled the project's reception bluntly: "It went extremely viral, so pretty fast we gained almost 20,000 GitHub stars" ^[3].

The project uses a planner-executor architecture: planner agents generate research questions, execution agents gather relevant information from across the web, and a publisher agent aggregates the findings into a structured report. An average research task takes roughly three minutes and costs about $0.10 in API calls. GPT Researcher supports multiple output formats, including PDF, Word, and Markdown, and can work with various LLM providers such as OpenAI, Claude, Gemini, Ollama, and Mistral ^[9].

GPT Researcher gained significant traction in the developer community. As of mid-2026, the project has accumulated more than 28,000 GitHub stars and over 3,800 forks. It is licensed under the Apache 2.0 license ^[9].

Company Formation

The viral success of GPT Researcher led Elovic to team up with Rotem Weiss, a data scientist and Columbia University graduate, to commercialize the underlying technology. In late 2023, they founded Tavily with the goal of building a search engine purpose-built for AI agents and LLM applications ^[3][10].

Yuval Rozio, Managing Director at Alpha Wave Global, also played a role in the company's formation, as Alpha Wave Global incubated the startup during its early stages ^[10].

Funding

Tavily raised about $5 million before securing its Series A. In August 2025, Insight Partners led a $20 million Series A funding round, with participation from Alpha Wave Global, bringing the company's total funding to $25 million ^[3][21]. At the time of the Series A, Tavily was roughly one year old and already generating revenue from a mix of AI startups and Fortune 500 enterprises ^[3].

By the time of the raise, Tavily reported more than 700,000 users, over 1 million monthly downloads, and roughly 100,000 GitHub mentions, growth the company said it achieved with effectively zero marketing spend through developer-led, organic adoption ^[21][22]. Insight Partners Managing Director George Mathew framed the opportunity around enterprise trust and compliance: "Governance, risk, and compliance at the enterprise is so important now, and if you just let that happen, it's just going to be the wild, wild west" ^[3]. Yuval Rozio of Alpha Wave Global, who helped incubate the company, said, "In just 12 months, the company grew into a category leader, proof that they're solving a real and urgent problem" ^[21].

Tavily said it would use the funding to double its headcount, deepen partnerships across the developer and LLM ecosystems, and expand precision retrieval of both internal and external data for developers worldwide ^[22].

Acquisition by Nebius

On February 10, 2026, Nebius Group N.V., an Amsterdam-based AI infrastructure company, announced an agreement to acquire Tavily ^[4][5]. Nebius is the international business that separated from Russia's Yandex: Yandex N.V. divested its Russian operations in July 2024 and was renamed Nebius Group N.V. that August, resuming trading on the NASDAQ under the ticker NBIS in October 2024 ^[4][23]. Bloomberg reported the initial deal value at $275 million in cash, with the total potentially rising to $400 million if certain performance milestones are met ^[4][5].

At the time of the announcement, Tavily reported 3 million monthly SDK downloads and a developer community of more than 1 million, growth Nebius attributed to organic, developer-led adoption ^[20]. Nebius operates data centers providing cloud compute for AI workloads and sought to add agentic search to its platform so customers would not need to manage a separate search vendor. Roman Chernin, Nebius co-founder and Chief Business Officer, said the acquisition "brings the search layer directly into our stack, so developers can focus on their applications instead of managing multiple vendors" and that Nebius is "building the complete platform for anyone who wants to build AI products, agents, or services" ^[5][20]. Tavily founder and CEO Rotem Weiss, who is joining Nebius along with his team, called agentic search "a multi-billion-dollar opportunity poised to grow exponentially" ^[20].

The acquisition occurred in the context of rapid growth in the agentic AI market, which is projected to expand from approximately $7 billion in 2025 to between $140 billion and $200 billion by the early 2030s ^[5].

Products and APIs

Tavily offers a suite of APIs designed to give AI agents and RAG systems structured access to real-time web data. All endpoints are authenticated using API keys and return results in JSON format optimized for LLM consumption ^[1][2].

Search API

The Search API is Tavily's core product. It performs semantic search optimized for AI workflows, returning concise, ranked snippets designed for downstream reasoning and retrieval-augmented generation. Unlike traditional search APIs that return raw URLs and brief snippets, Tavily aggregates up to 20 sites per request, uses proprietary AI to score and filter sources, and returns structured results with titles, content snippets, source URLs, and relevance scores ^[1][11].

The Search API supports four depth levels that trade off between latency and relevance:

Scroll sideways for more →

The API also supports topic categories (general, news, and finance), temporal filtering (by day, week, month, year, or specific date ranges), domain inclusion and exclusion lists, country-based result boosting, safe search filtering, and an auto-parameters mode that automatically configures search settings based on query intent ^[11].

When the include_answer parameter is enabled, Tavily generates a short LLM-powered answer alongside the search results, which is useful for agent-to-agent communication. The include_raw_content option returns the full parsed content of each page in Markdown or plain text format ^[11].

Extract API

The Extract API retrieves raw content from a list of specified URLs, making it suitable for data collection, content analysis, and research tasks. It converts web pages into clean, structured text that LLMs can process directly, and a single call can process up to 20 URLs ^[2][12]. The API supports two extraction depth levels: basic and advanced. Users can choose between Markdown and plain text output formats, and can optionally include images and favicons in the response ^[12].

Extract costs 1 credit per 5 successful URL extractions at basic depth, or 2 credits per 5 URLs at advanced depth ^[13].

Map API

The Map API discovers and visualizes the structure of a website starting from a base URL. It returns a list of pages and their relationships, allowing developers to understand a site's layout before extracting content. Map costs 1 credit per 10 pages discovered, or 2 credits per 10 pages when custom instructions are provided ^[12][13].

Crawl API

The Crawl API automatically explores and extracts content from an entire website by providing a single starting URL. It traverses a site like a graph, following internal links and gathering content from each page it visits. Developers can control the crawl's depth, breadth, and total page limit. Additional parameters include natural language queries to prioritize specific content, regex-based path and domain filters, the option to follow external links, and category filters for targeting specific page types such as documentation or blog posts ^[14].

Crawl combines the costs of mapping and extraction, so the total credit usage depends on the number of pages discovered and the extraction depth selected ^[13].

Research API

The Research API provides an automated deep research capability built on top of the other APIs. It generates comprehensive research reports on any topic by coordinating multiple search, extraction, and synthesis steps. Tavily describes the Research endpoint as an "Agent-in-a-Box" that performs multiple iterative searches, reasons over the retrieved data, deduplicates sources, and returns structured JSON in a single call, and the company says the endpoint outperforms in-house deep-research implementations from OpenAI and Perplexity on the DeepResearch Bench evaluation ^[2]. Two model tiers are available: the "pro" model costs between 15 and 250 credits per request, while the "mini" model costs between 4 and 110 credits per request ^[13].

Technical Architecture

Search Pipeline

Tavily's search pipeline is designed to transform raw web data into LLM-ready content in a single API call. When a query is submitted, the system performs the following steps ^[1][2]:

1. The query is processed and, if auto-parameters are enabled, the system automatically selects optimal search parameters based on query intent.
2. Multiple web sources are searched simultaneously, aggregating results from up to 20 sites.
3. Proprietary AI models score, filter, and rank the results based on relevance to the query.
4. Content is extracted, cleaned, and converted into structured snippets suitable for LLM processing.
5. Results are returned in a standardized JSON format with relevance scores, source URLs, and optional raw content.

The architecture incorporates dynamic caching and an agent-native index to maintain low latency even at production scale ^[2].

Security Layer

Tavily positions itself, in its own words, as "a firewall between your agents and the internet" ^[2]. All retrieved content passes through security, privacy, and content validation layers that are designed to block several categories of threats ^[2][15]:

• Prompt injection attempts embedded in web content
• Personally identifiable information (PII) leakage
• Malicious or manipulated source content
• Adult or unsafe content (with safe search filtering for enterprise customers)

Tavily holds SOC 2 certification and operates with a zero data retention policy, meaning that search queries and results are not stored after processing. The company also partnered with Pillar Security to provide additional enterprise-grade security guardrails for AI web access ^[15].

SDKs and Developer Tools

Python SDK

The Tavily Python SDK (tavily-python) provides a full-featured client library for interacting with all Tavily API endpoints. It supports synchronous and asynchronous operations (async support was added in version 0.3.4), and covers the complete range of search, extract, crawl, map, and research functionalities. The package is available on PyPI ^[16].

JavaScript/TypeScript SDK

The official JavaScript SDK (@tavily/core) offers the same capabilities as the Python SDK for Node.js and browser environments. A separate package (@tavily/ai-sdk) provides compatibility with Vercel's AI SDK (versions 5 and 6) for building AI-powered web applications ^[16].

REST API

All Tavily functionality is accessible through a RESTful API at https://api.tavily.com/. Authentication is handled via Bearer tokens in the HTTP Authorization header. The API returns JSON responses and supports standard HTTP error codes for invalid requests (400), authentication failures (401), rate limiting (429), plan limit exceedance (432), and server errors (500) ^[11].

MCP Server

Tavily provides a Model Context Protocol (MCP) server that connects AI assistants such as Claude, Cursor, and other MCP-compatible tools directly to Tavily's search, extract, map, and crawl capabilities. The MCP server is available both as a remote hosted service (at https://mcp.tavily.com/mcp/) and as a locally installable package. Authentication can be handled through OAuth or API key query parameters ^[8].

Integrations with AI Frameworks

LangChain

Tavily has a dedicated LangChain integration package (langchain-tavily for Python, @langchain/tavily for JavaScript). This integration provides tool classes that can be bound directly to LangChain agents, giving them the ability to perform real-time web searches. The tool accepts customizable parameters including max_results, search_depth, topic, time_range, include_domains, exclude_domains, and more. Tavily is one of the default search tools recommended in LangChain's agent documentation ^[7].

LlamaIndex

The TavilyToolSpec for LlamaIndex allows developers to integrate Tavily's search and research capabilities into LlamaIndex agent workflows. The integration is available through the llama-index-tools-tavily-research package on PyPI and is listed on LlamaHub ^[17].

CrewAI

While CrewAI does not include a native Tavily integration out of the box, any LangChain or LlamaIndex tool can be converted into a CrewAI tool. This allows CrewAI multi-agent systems to use Tavily for web search and content extraction tasks ^[18].

Vercel AI SDK

The @tavily/ai-sdk package provides direct compatibility with Vercel's AI SDK, enabling developers building Next.js or other React-based AI applications to integrate real-time web search with minimal setup ^[16].

Other integrations

Beyond the major frameworks, Tavily documents integrations with agent tooling such as Agno and the low-code automation platform n8n, alongside its native SDKs and MCP server ^[2].

Pricing

Tavily offers a tiered pricing structure with monthly subscription plans and a pay-as-you-go option. All plans include access to the Search, Extract, Map, and Crawl APIs. Credits do not roll over between billing periods ^[13].

Scroll sideways for more →

The free Researcher plan provides 1,000 API credits per month with no credit card required, making it accessible for individual developers and small experiments. Enterprise plans offer custom credit volumes, dedicated support, and enhanced security features ^[13].

Credit Costs by Operation

Scroll sideways for more →

Use Cases

Retrieval-Augmented Generation

Tavily's primary use case is providing the retrieval layer for RAG systems. Instead of building and maintaining custom web scraping infrastructure, developers can use the Search API to fetch relevant, up-to-date information that supplements the knowledge of their LLMs. The structured JSON output is designed to be inserted directly into LLM prompts as context, reducing the need for post-processing ^[1][2].

AI Agent Web Access

For autonomous AI agents, Tavily serves as the interface between the agent and the live web. Agents can call the Search API to answer factual questions, the Extract API to read specific web pages, and the Crawl API to index entire documentation sites. This is particularly valuable for agents performing tasks like lead research in sales, fraud detection in finance, and customer support automation ^[3][6].

Automated Research

The Research API and the open-source GPT Researcher project enable fully automated research workflows. Given a topic, the system generates research questions, searches the web for relevant information, synthesizes findings, and produces a multi-page report with citations. This approach is used for market research, competitive analysis, academic literature reviews, and content creation ^[9].

Knowledge Base Construction

The Crawl API enables organizations to build and maintain knowledge bases by systematically extracting content from documentation sites, wikis, and other structured web sources. Extracted content can be stored in vector databases for semantic search or used to populate internal knowledge management systems ^[14].

Chatbot Grounding

AI chatbots use Tavily to ground their responses in current, verifiable information. By calling the Search API before generating a response, chatbots can provide answers that include source citations and reflect the latest available information rather than relying solely on their training data ^[1].

Competition

Tavily operates in the growing market for AI-native search APIs. How does Tavily compare to alternatives such as Exa and SerpAPI? Several competitors offer overlapping capabilities with different strengths and trade-offs, summarized below.

Scroll sideways for more →

Other competitors include WebSearchAPI, Firecrawl (focused on web scraping and crawling), Perplexity's Sonar API (focused on conversational search), and Serper (a lightweight Google SERP API) ^[19].

GPT Researcher

GPT Researcher remains Tavily's flagship open-source project and serves as both a showcase for Tavily's API capabilities and a standalone research tool. The project is maintained at github.com/assafelovic/gpt-researcher under the Apache 2.0 license and, as of mid-2026, has more than 28,000 GitHub stars and over 3,800 forks ^[9].

Architecture

GPT Researcher uses a multi-agent architecture inspired by the STORM paper from Stanford University. The system coordinates three types of agents ^[9]:

• Planner agents analyze the research topic and generate a set of targeted research questions.
• Execution agents search the web using Tavily's API (by default) to gather relevant information for each question.
• Publisher agents synthesize the collected information into a comprehensive, cited research report.

Capabilities

As of early 2026, GPT Researcher offers the following features ^[9]:

• Generates research reports exceeding 2,000 words with proper citations
• Aggregates information from 20 or more sources per report
• Supports deep research mode with recursive, tree-like exploration (roughly 5 minutes, about $0.40 per research task)
• Exports to PDF, Word, and Markdown formats
• Includes smart image scraping and AI-generated illustrations via Google Gemini
• Supports JavaScript-enabled web scraping
• Integrates with MCP for connecting to specialized data sources
• Works with local documents (PDF, Excel, CSV, Word, Markdown, PowerPoint)
• Supports multiple LLM providers including OpenAI, Claude, Gemini, Ollama, and Mistral
• Provides LangSmith tracing support for observability

Technology Stack

The backend is built with Python and FastAPI, with LangGraph support for multi-agent workflows. The frontend is available as either a lightweight HTML/CSS/JavaScript interface or a production-grade Next.js application with Tailwind CSS. The default retrieval backend uses Tavily's API, though other search providers can be configured ^[9].

Is Tavily open source?

Tavily's commercial platform, its Search, Extract, Map, Crawl, and Research APIs, is a proprietary, paid product accessed through API keys and metered in credits ^[13]. Its origins and much of its developer following, however, come from open source. GPT Researcher, the autonomous research agent that Assaf Elovic released in 2023, is distributed under the permissive Apache 2.0 license and, as of mid-2026, has more than 28,000 GitHub stars ^[9]. Tavily also publishes open-source client SDKs for Python and JavaScript/TypeScript and an open Model Context Protocol server, so developers can integrate the service freely even though the underlying search index and ranking models remain closed ^[8][16]. In short, the tooling around Tavily is open, while the search service itself is a commercial API.

Company Overview

Scroll sideways for more →

See Also

• Retrieval-augmented generation
• AI agent
• LangChain
• LlamaIndex
• Information retrieval
• Vector database
• Model Context Protocol

References

1. "About." Tavily Documentation. https://docs.tavily.com/documentation/about ↩
2. "Tavily 101: AI-powered Search for Developers." Tavily Blog. https://www.tavily.com/blog/tavily-101-ai-powered-search-for-developers ↩
3. Wiggers, Kyle. "Tavily raises $25M to connect AI agents to the web." TechCrunch, August 6, 2025. https://techcrunch.com/2025/08/06/tavily-raises-25m-to-connect-ai-agents-to-the-web/ ↩
4. "Nebius Agrees to Buy AI Agent Search Company Tavily for $275 Million." Bloomberg, February 10, 2026. https://www.bloomberg.com/news/articles/2026-02-10/nebius-agrees-to-buy-ai-agent-search-company-tavily-for-275-million ↩
5. "AI infrastructure giant Nebius buys agentic search startup Tavily." SiliconANGLE, February 10, 2026. https://siliconangle.com/2026/02/10/ai-infrastructure-giant-nebius-buys-agentic-search-startup-tavily/ ↩
6. "Enterprise Solutions." Tavily. https://www.tavily.com/enterprise ↩
7. "Tavily search integration." LangChain Documentation. https://docs.langchain.com/oss/javascript/integrations/tools/tavily_search ↩
8. "Tavily MCP Server." Tavily Documentation. https://docs.tavily.com/documentation/mcp ↩
9. Elovic, Assaf. "GPT Researcher." GitHub. https://github.com/assafelovic/gpt-researcher ↩
10. "Tavily." Crunchbase. https://www.crunchbase.com/organization/tavily ↩
11. "Search API Reference." Tavily Documentation. https://docs.tavily.com/documentation/api-reference/endpoint/search ↩
12. "API Reference Introduction." Tavily Documentation. https://docs.tavily.com/documentation/api-reference/introduction ↩
13. "Credits & Pricing." Tavily Documentation. https://docs.tavily.com/documentation/api-credits ↩
14. "What is Tavily Crawl API?" Tavily Help Center. https://help.tavily.com/articles/5815909991-what-is-tavily-crawl-api ↩
15. "Tavily Announces Strategic Partnership with Pillar to Secure AI Web Access." Tavily Blog. https://blog.tavily.com/tavily-partners-with-pillar-to-deliver-enterprise-grade-ai-web-access-with-built-in-security-to-secure-ai-agents-web-access-2/ ↩
16. "Tavily Python SDK." GitHub. https://github.com/tavily-ai/tavily-python ↩
17. "Tavily Research Tool." LlamaHub. https://llamahub.ai/l/tools/llama-index-tools-tavily-research ↩
18. "LlamaIndex." Tavily Documentation. https://docs.tavily.com/documentation/integrations/llamaindex ↩
19. "Best SERP API Comparison 2025." DEV Community. https://dev.to/ritza/best-serp-api-comparison-2025-serpapi-vs-exa-vs-tavily-vs-scrapingdog-vs-scrapingbee-2jci ↩
20. "Nebius announces agreement to acquire Tavily to add agentic search to its AI cloud platform." Nebius Newsroom, February 10, 2026. https://nebius.com/newsroom/nebius-announces-agreement-to-acquire-tavily-to-add-agentic-search-to-its-ai-cloud-platform ↩
21. "Tavily Raises $25 Million to Power the Internet of Agents." Insight Partners, August 6, 2025. https://www.insightpartners.com/ideas/tavily-raises-25-million-to-power-the-internet-of-agents/ ↩
22. "Big News from Tavily: Announcing 25M Series A to Power the Internet of Agents." Tavily Blog, August 6, 2025. https://www.tavily.com/blog/big-news-from-tavily-announcing-25m-series-a-to-power-the-internet-of-agents ↩
23. "Nebius Group N.V. announces official name change and new ticker symbol." Nebius Newsroom, August 2024. https://nebius.com/newsroom/nebius-group-n-v-announces-official-name-change-and-new-ticker-symbol ↩
24. "Connect your AI agents to the web." Tavily Product Page. https://www.tavily.com/product ↩