docs: align English copy with current product names and positioning

[erikengervall]

Updates English pages and shared snippets where the copy still uses retired names, old package names, or dated taglines. Localized copies are left to the translation pipeline.

Package names and links

• Python installs now use pip install firecrawl (quickstarts, advanced scraping guide, capabilities page, and the shared v2 install and proxy-retry snippets that feed the SDK and feature pages). firecrawl-py keeps working as an alias.
• JS imports in the Search page, the web-scale monitoring page, and the v2 Monitor snippets now import from firecrawl instead of @mendable/firecrawl-js.
• The v2 crawl WebSocket sample crawls https://firecrawl.dev instead of https://mendable.ai, and the v2 search sample output uses the firecrawl/firecrawl repo title.

Retired names

• Agent page: Agent is described as replacing the legacy /extract endpoint, and the comparison table is labelled accordingly.
• Search page: drops the FIRE-1 mention from the scrape-options note (v2 scrape options have no agent option).
• Billing: the FIRE-1 row is renamed to the v1 scrape agent option it actually bills.
• Product lists that named "extract" as a capability now name Agent where the surface has it (CLI on AI onboarding and OpenClaw, Composio, the Stripe Projects endpoint list), and the Hermes card says scrape. The Lovable table describes structured data as Scrape's JSON format, since the Lovable connector has no extract action.

Taglines and descriptions

• Introduction, LLM SDK guides (Claude, Gemini, LangChain, OpenAI), SDK page descriptions, Express/NestJS/ASP.NET Core quickstarts, SourceSync, Enterprise, and a few use-case pages drop "web data API", "turn websites into markdown", and "LLM-ready data" phrasing in favour of the current wording (the context API for AI agents; search, scrape, and interact; LLM-ready context).
• Scrape and Parse descriptions and sample outputs updated to match.
• The "Ready to build?" sidebar card now reads "Search, scrape, and interact with the web for free".
• Observability use case points to Monitor for scheduled change detection before describing the do-it-yourself scrape pipeline.
• ADK guide: FIRECRAWL_API_KEY is documented as optional with the cloud API, matching the MCP server's keyless mode for scrape and search.
• Brand casing fix in one Learn tutorial comment.

Left as is on purpose

• LangChain, LlamaIndex, and CrewAI install lines keep @mendable/firecrawl-js / firecrawl-py, because those packages declare the alias as their dependency.
• Zapier, Make, Pipedream, LangChain, and LlamaIndex pages keep "extract" where it is the third-party action or mode name.
• Webhook event types, Lockdown availability, SIEM event sources, the Ask job list, and the key-restrictions v0 note still mention extract or v0, because those endpoints still answer and the text documents their behaviour.
• Keyless wording on the MCP, rate-limits, Developer, and AI onboarding pages already matches the API's per-day keyless tier, so it is unchanged.

Needs a human look

• Sidebar CTA wording (ready-to-build-cta.js) shows on every page.
• Billing row label "v1 scrape agent option".
• Composio row swapped from Extract to Agent, and Lovable row relabelled "Structured data".
• Sample outputs (search titles, scrape ogDescription, company_mission) use current taglines, not a captured response.
• Use-case and LLM-guide one-liners (overview, AI platforms, developers & MCP, observability, Enterprise intro, SourceSync).
• The Observer template cards on the competitive-intelligence, investment-finance, and observability pages still point to the firecrawl-observer repo; consider swapping them for Monitor.

Checks: scripts/check-locale-api-literals-selftest.sh and scripts/check-extraction-hostile-markdown.sh pass locally.

[mintlify]

Preview deployment for your docs. Learn more about Mintlify Previews.

Project	Status	Preview	Updated
firecrawl	🟢 Ready	View Preview	Oct 9, 2026, 2:30 AM

💡 Tip: Enable Automations to automatically generate PRs for you.

[capy-ai]

Closing in favour of smaller PRs, one change each, all branched from main so they can merge independently. Together they carry exactly this diff:

• docs: install and import the firecrawl package #1568 Install and import the firecrawl package
• docs: replace mendable.ai and mendableai in v2 samples #1567 Replace mendable.ai and mendableai in v2 samples
• docs: drop the FIRE-1 name from Billing and Search #1570 Drop the FIRE-1 name from Billing and Search
• docs: describe Agent as replacing the legacy /extract #1564 Describe Agent as replacing the legacy /extract
• docs: list Agent instead of extract in capability lists #1569 List Agent instead of extract in capability lists
• docs: use the context API tagline on the Introduction #1566 Context API tagline on the Introduction
• docs: update SDK page descriptions #1565 SDK page descriptions
• docs: update LLM framework guide intros #1563 LLM framework guide intros
• docs: update Express, NestJS, and ASP.NET Core quickstart descriptions #1571 Express, NestJS, and ASP.NET Core quickstart descriptions
• docs: update use-case intro lines #1573 Use-case intro lines
• docs: name the products in the Enterprise intro #1572 Enterprise intro
• docs: update the SourceSync.ai description #1574 SourceSync.ai description
• docs: update Scrape and Parse descriptions #1575 Scrape and Parse descriptions
• docs: use current taglines in sample outputs #1576 Current taglines in sample outputs
• docs: update the Ready to build sidebar card #1577 "Ready to build?" sidebar card
• docs: point the observability use case to Monitor #1579 Observability use case points to Monitor
• docs: document the ADK API key as optional on the cloud API #1578 ADK API key documented as optional on the cloud API
• docs: fix Firecrawl casing in the RAG Llama 3 tutorial #1580 Firecrawl casing in the RAG Llama 3 tutorial