Firecrawl, the web data platform for AI agents, released a series of updates between March and October 2026 that expand what automated tools can retrieve from the web. The changes include a new knowledge library called Alexandria, a $75M Series B funding round, specialized search indexes for developers and researchers, and lower token costs for extracting specific data from pages.
The most consequential shift is Alexandria, introduced September 22, 2026. It consolidates live web search, official data providers, site-specific connectors, and Firecrawl's own indexes into a single interface that AI agents can query. The company said agents using Alexandria scored 21% higher on answer quality than those using built-in web tools, based on blind AI judging across 845 tasks with the same model and prompts.
What Alexandria unlocks for agent workflows
Alexandria gives agents access to more than 100 data providers alongside Firecrawl's existing search and scraping capabilities. Agents can discover a source, inspect its inputs and price, and retrieve data through the same Firecrawl connection. Discovery and inspection are free; execution is billed at each tool's listed price.
In October, Firecrawl added people and company enrichment through Apollo, FullEnrich, and Data Legion. Agents can find decision makers by title, company domain, and location, then enrich profiles with work emails and employment history. Company enrichment covers funding, headcount, technologies, job postings, and news. Some providers require an organization admin to accept their terms before the first call.
Alexandria is available in Claude, Claude Code, ChatGPT, and Codex through Firecrawl plugins, as well as through the Firecrawl API, MCP, and CLI. The company's example prompt: "Use Firecrawl Alexandria to find engineering leaders at a company I specify. Show their current roles and the available work email enrichment options with prices before running enrichment."
Specialized indexes for research and development
Firecrawl launched its Research Index in June, covering more than 3 million arXiv papers plus GitHub artifacts from top research repos. In August, the company added a Life Sciences category with 41 million papers across drug discovery, clinical trial, and biology literature. The index is free to use and hits 90% recall@10 on Firecrawl's paper-retrieval evaluation.
The Developer Index, released August 20, targets coding agents with 70 million artifacts across READMEs, documentation, issues, pull requests, and OpenAPI specs. Firecrawl reports a 0.63 recall@10 on its open DevDex benchmark, which it says is the highest of any major coding-specific index tested. Every result carries a stable ID, URL, and matched passages in markdown.
For professionals who don't write code, the practical takeaway is that AI assistants connected to these indexes can answer questions about products, companies, and technical documentation with citations to primary sources rather than general web pages.
Token efficiency and monitoring
Several 2026 releases focus on reducing the cost of using AI agents on web data. The Highlights format returns only the exact sentences, code blocks, and table rows that match a query, using up to 100x fewer tokens than processing full pages. The Question format returns grounded answers from a page without returning the page itself. Firecrawl's updated /search endpoint uses a custom relevance model and reports 94.7% accuracy on SimpleQA.
The /monitor endpoint, introduced in May and expanded in July, watches pages or entire search queries for changes and alerts agents via webhook or email. Users describe what to track in plain English, and the system configures URLs, schemas, and schedules. The company said the approach can reduce LLM token usage by up to 90% because agents only ingest what actually changed.
Why this matters for sales, HR, and operations teams
People enrichment through Alexandria is the most directly useful feature for sales and recruiting teams. You can build a target account list, find the people who own a relevant function, and look up work emails for a shortlist without switching between multiple data providers. HR teams can enrich inbound applications with company context. Marketing and operations teams can set up /monitor alerts to track competitor pricing pages, job boards, or regulatory sites and receive notifications only when something actually changes.
The pricing model is transparent: each tool shows its execution price before an agent calls it, and discovery is free. For teams already paying for web scraping or manual data enrichment, consolidating those workflows through one API connection may reduce both tooling costs and the time spent switching between platforms. If you're evaluating AI tools for business development, the AI Sales Strategy Training path covers how these agent-driven workflows fit into existing sales processes.
Your membership also unlocks:
