Tools · SPBoucher
Menu

Exa

by Exa Labs · API service · also known as Exa AI, Metaphor

Search API for AI agents built on Exa's own web index, offering search, page contents, deep research and asynchronous agent endpoints with Python and JavaScript SDKs

ProprietaryFreemiumUsage-basedEnterpriseCloud
Visit official site Compare

Link verified Nov 2, 2026

Overview

Exa (formerly Metaphor) is a web search API from Exa Labs designed for programmatic use by language-model applications. Unlike wrappers around consumer search engines, Exa operates its own crawl and index and exposes semantic as well as keyword retrieval. The Search endpoint returns results with optional page text, highlights or summaries trimmed to be token-efficient; Contents fetches full text for any URL; Deep Search reasons across sources and returns structured output; and asynchronous Agent endpoints handle research reports, list building and data enrichment.

Developers integrate through the REST API, the exa-py and exa-js SDKs (MIT-licensed), a hosted MCP server and an installable agent skill. Pricing is pay-as-you-go per request with a free monthly credit allowance and enterprise contracts for volume.

Exa is used for research assistants, lead and company discovery, and grounding agents in current web content. It is a proprietary hosted service whose index coverage and ranking are not inspectable, it cannot read paywalled content, and heavy use of content retrieval adds cost on top of search calls.

Capabilities

  • Semantic and keyword web search over a proprietary index
  • Full-text, highlights and summaries for result pages
  • Deep search with structured output
  • Asynchronous research and enrichment agents
  • Python and JavaScript SDKs
  • Hosted MCP server

Typical workflows

  1. Find similar companies or papers from an example URL
  2. Ground an assistant with fresh web sources
  3. Enrich a spreadsheet of entities with web data

Strengths

  • Independent index built for machine consumers
  • Semantic retrieval beyond keyword matching
  • No subscription; pay per request

Limitations

  • Proprietary hosted service
  • Index coverage and ranking are opaque
  • Costs grow with content retrieval volume

Prerequisites

  • API key

Who it is for

  • Agent and RAG developers
  • Analysts automating research and discovery

Alternatives & comparison

  • Brave Search API

    Developer API over Brave's independent web index offering web, news, image, video, LLM-context and grounded-answer endpoints on prepaid credits

    ProprietaryFreemium
  • Jina Reader

    Prefix-style API from Jina AI that converts any public URL into LLM-ready Markdown (r.jina.ai) or returns web search results with content (s.jina.ai), with open-source server code

    MixedFreemium
  • SerpApi

    API that returns live search-engine result pages from Google and other engines as structured JSON, with localization, device and search-type parameters

    ProprietaryFreemium
  • Tavily

    Web search, extraction, crawling and research API designed for LLM agents, returning concise source content with SDKs, framework integrations and a hosted MCP server

    ProprietaryFreemium
Compare Exa with its alternatives →

Sources & verification

Primary sources consulted for this record. Claims are tied to the fields each source supports; the automated link checker re-verifies every URL on a schedule.

  • Official documentation
    Exa documentation

    https://exa.ai/docs

    Supports: capabilities, languages, description

    checked Nov 1, 2026
  • Official site
    Exa | Search API for AI Agents

    https://exa.ai/

    Supports: name, summary

    checked Nov 1, 2026
  • Pricing page
    Exa pricing

    https://exa.ai/pricing

    Supports: pricing_model

    checked Nov 1, 2026
  • Repository
    exa-labs/exa-py

    https://github.com/exa-labs/exa-py

    Supports: vendor, languages

    checked Nov 1, 2026
Confidence
High
Reviewed
Nov 1, 2026
Links
Verified · Nov 2, 2026

Spotted an error?

Broken link, outdated license, wrong category, or a tool that no longer exists? Tell us; reports are reviewed by the editor.

Firecrawl

Firecrawl (formerly Mendable) · Data collection & crawling

In-depth entry

Web data API that scrapes, crawls, maps and searches websites and returns clean Markdown or structured JSON for LLM pipelines, with an AGPL-licensed self-hostable core

AGPL-3.0-onlyFreemiumPaidSelf-hostedNov 2, 2026

Bright Data

Bright Data Ltd. · Data collection & crawling

Web data platform combining residential, datacenter and ISP proxy networks with unblocking, SERP and scraper APIs, remote browsers, ready-made datasets and an MCP server

ProprietaryFreemiumUsage-basedNov 2, 2026

Jina Reader

Jina AI · Data collection & crawling

Prefix-style API from Jina AI that converts any public URL into LLM-ready Markdown (r.jina.ai) or returns web search results with content (s.jina.ai), with open-source server code

Apache-2.0FreemiumUsage-basedSelf-hostedNov 2, 2026

Meilisearch

Meili SAS · Vector databases

Rust search engine for application search-as-you-type with typo tolerance, faceting and hybrid semantic search; MIT core plus BSL enterprise parts

MIT AND BUSL-1.1Open sourceFreeSelf-hostedNov 2, 2026

Typesense

Typesense, Inc. · Vector databases

GPL-licensed in-memory search engine in C++ offering typo-tolerant instant search, faceting, vector and hybrid search, with Typesense Cloud hosting

GPL-3.0Open sourceFreeSelf-hostedNov 2, 2026