Firecrawl Release Notes

• Date: March 25, 2026

Scrape any page, then let your agent dive deeper into what it finds. The new /interact endpoint turns any scrape into a live browser session where your agent can click, type, and navigate - all in natural language.

Interact CLI

Highlights

• Natural Language Control - Describe what you want in plain English. The agent clicks, types, scrolls, and extracts data automatically. No selectors, no scripts.
• Live Browser Sessions - Every session includes a live URL you can embed, share, or interact with in real time. Build browser-powered UIs, debug visually, or demo workflows.
• Persistent Profiles - Log in once and pick up wherever you left off. Cookies, localStorage, and session state carry across scrapes with named profiles - no re-authenticating, no lost state.
• Full Playwright Control - Need precision? Switch to code mode and run Playwright (Node.js or Python) or Bash with agent-browser in the same session.
• Session Reuse - Chain multiple interact calls on the same scrape. The browser stays open and maintains state between calls, so complex multi-step workflows just work.

View the full documentation here.

• Date: February 26, 2026

Our new PDF parsing engine delivers 3x faster parsing and significantly improved reliability.

Rebuilt in Rust, it automatically adapts to any PDF - from clean text files to scanned reports and complex layouts - ensuring agents receive complete, accurate data every time.

PDF Parser V2

Highlights

• Rust-Based Parser - A new high-performance engine built in Rust delivers up to 3x faster parsing, dramatically reducing latency in data ingestion and embedding workflows.
• Three Parsing Modes - Choose the right balance of speed and completeness:
  • fast - text-only parsing with the Rust engine for maximum performance.
  • auto - the new default; starts in fast mode and automatically falls back to OCR when needed. Auto intelligently detects when PDFs require OCR even if they appear text-based, handling edge cases like embedded images, graphs, multi-column layouts, and unusual text encodings.
  • ocr - forces OCR parsing for fully image-based or scanned documents.
• Built for Production Reliability - Extensively tested across thousands of real-world PDFs to ensure consistent, accurate extraction in production environments.

Read the full blog here.

• Date: February 17, 2026

Browser Sandbox gives your agents a secure, fully managed browser environment for interactive web automation. No local setup, no Chromium installs, no driver compatibility issues. Each session runs in an isolated, disposable sandbox that scales without managing infrastructure.

Browser Sandbox

Highlights

• Browser Sandbox - Launch secure, isolated browser sessions with Python, JavaScript, and bash execution. Pre-installed with agent-browser CLI and Playwright.
• Multi-Language Support - Execute Python, JavaScript, or bash code remotely. Send commands via API, CLI, or SDK and get results instantly.
• agent-browser Integration - Pre-installed CLI with 40+ commands. AI agents write simple bash commands instead of complex Playwright code.
• Live View & CDP Access - Watch sessions in real time via embeddable stream URL or connect your own Playwright instance over WebSocket.
• Session Management - Configurable TTL controls, parallel sessions (up to 20 concurrent), and automatic cleanup. 2 credits per browser minute.

Read the full documentation here.

• Date: February 6, 2026

Branding Format v2

We significantly improved logo extraction accuracy for Branding Format v2, our endpoint for extracting brand identities from websites.

Highlights

• Significantly improved logo detection: Logo extraction is now far more reliable, with fewer false positives and better handling of edge cases like logos embedded in background images.
• Works with modern site builders: Branding Format now properly detects logos built with Wix, Framer, and other drag-and-drop platforms that generate complex or non-semantic HTML.
• Built for AI agents and developers: Branding Format goes beyond logos - capturing colors, typography, spacing, and UI components in a structured format to power AI agents and apps.

Read the full blog here.

• Date: February 3, 2026

v2.8.0 Launch

Firecrawl v2.8.0 brings major improvements to agent workflows, developer tooling, and self-hosted deployments across the API and SDKs.

Highlights

• Parallel Agents - Execute thousands of /agent queries simultaneously with automatic failure handling and intelligent waterfall execution. Powered by Spark 1 Fast for instant retrieval, automatically upgrading to Spark 1 Mini for complex queries requiring deeper research.
• Firecrawl Skill - Enables agents to use Firecrawl for web scraping and data extraction, install via npx -y firecrawl-cli@latest init --all --browser.
• Firecrawl CLI - Command-line interface with full scrape, search, crawl & map support, install via npm install -g firecrawl-cli.
• Spark Model Family - Three new models powering /agent: Spark 1 Fast for instant retrieval (currently only available in Playground), Spark 1 Mini for complex research queries, and Spark 1 Pro for advanced extraction tasks.
• Agent Enhancements - Webhook support, model selection, and new MCP Server tools for autonomous web data gathering.

Read the full changelog here.

• Date: January 30, 2026

We're bringing parallel processing to /agent, letting you batch hundreds or even thousands of queries simultaneously. What took hours of sequential agent queries now completes in minutes with automatic failure handling and parallel execution.

Try it in the Agent Playground - the playground automatically determines when your use case is a good fit for Parallel Agents.

Parallel Agents

Highlights

• Parallel Batch Processing: Run thousands of /agent queries simultaneously to enrich companies, research competitors, or build product datasets at scale.
• Intelligent Waterfall: Tries instant retrieval first, then automatically upgrades specific cells to full agent research (Spark One Mini) only when needed.
• Real-Time Spreadsheet Interface: Work in familiar CSV format with instant visual feedback as cells populate in real-time.
• Zero Configuration: Input your data schema, write one prompt, hit run. No workflow building required.
• Predictable Pricing: 10 credits per cell with Spark-1 Fast.

Read the full blog here.

• Date: January 27, 2026

Firecrawl Skill + CLI We're introducing the Firecrawl Skill and CLI, a new way for AI agents to reliably access real-time web data on their own. With a single install, agents like Claude Code, Antigravity, and OpenCode can access all of your favorite Firecrawl endpoints - including scrape, search, crawl, and map for any use case you need.

Install with $ npx -y firecrawl-cli@latest init --all --browser and learn more in our docs.

Highlights

• One-Command Install: Install the skill with a single command and teach agents how to install, authenticate, and use all of Firecrawl's powerful end-to-end.
• Real-Time Web Data at Runtime: Agents can pull fresh, full-page content from docs, product pages, pricing, and articles exactly when they need it.
• Context-Efficient for Agents: Uses a file-based approach for context management and bash methods for efficient search and retrieval.
• Works Across Complex & Dynamic Sites: Powered by Firecrawl's custom browser stack to reliably extract complete data from large, JavaScript-heavy sites.
• Proven, Best-in-Class Coverage: Backed by benchmark results showing >80% coverage across real-world evaluations.

Read the full blog here.

• Date: December 5, 2025

v2.7.0 Launch

Highlights

• ZDR Search Support - Enterprise customers can now search with Zero Data Retention enabled end to end. If you're interested, contact alex@firecrawl.dev to enable for your team.
• Partner Integrations API - Available in closed beta for native integrations. Get in touch with us at partnerships@firecrawl.dev if you are intested in offering Firecrawl as a native integration in your product.
• Improved Branding Format - Better detection and support across all platforms.
• Faster Screenshots - Enhanced viewport and full page screenshots with improved speed and accuracy.
• Self-hosted Improvements - Significant enhancements for deployments and infrastructure.
• Performance Enhancements - Platform-wide improvements for better user experience.

Read the full changelog here

• Date: November 14, 2025

v2.6.0 Launch

Highlights

• Unified Billing Model - Credits and tokens merged into single system. Extract now uses credits (15 tokens = 1 credit), existing tokens work everywhere.
• Enhanced Branding Format - Full support across Playground, MCP, JS and Python SDKs.
• Reliability and Speed Improvements - All endpoints significantly faster with improved reliability.
• Instant Credit Purchases - Buy credit packs directly from dashboard without waiting for auto-recharge.
• Improved Markdown Parsing - Enhanced markdown conversion and main content extraction accuracy.
• Change Tracking - Faster and more reliable detection of web page content updates.
• Core Stability Fixes - Fixed tons of core stability issues, PDF timeouts, and improved error handling.

Read the full changelog here

• Date: October 25, 2025

v2.5.0 Launch

Today, we're excited to announce Firecrawl v2.5, which delivers the highest quality and most comprehensive web data API available. This release represents a significant leap forward in web data extraction, powered by two major infrastructure improvements: our new Semantic Index and a completely custom browser stack.

See the benchmarks below:

Benchmarks

We've open-sourced these benchmarks! Check out scrape-evals, our reproducible framework for testing web scraping engines on 1,000 real URLs.

Highlights

• Open-Source Scrape-Evals Benchmark
  We've released scrape-evals, an open-source benchmark testing 13 web scraping engines on 1,000 real URLs for coverage and quality.

• Full-Page, High-Quality Extraction
  Improved browser stack ensures complete and consistent data from any type of website.

• Semantic Index for Faster Results
  Retrieve either fresh data or a previously indexed snapshot with faster speeds and increased coverage.

• 5x Cheaper Search & New Credit Packs
  Search is now 5x cheaper and now every plan has an auto-recharge credit pack sized to match your scale.

• Smarter Concurrency & Crawl Architecture
  New crawling system improves throughput, reliability, and queue fairness across large workloads. …

• Date: October 13, 2025

v2.4.0 Launch

What's New

• New PDF Search Category - Search specifically for PDF documents using our v2/search endpoint with the new 'pdf' category filter
• 10x Better Semantic Crawling - Improved accuracy and relevance when crawling with a prompt
• New x402 Search Endpoint - Our search API available via Coinbase x402 integration
• Fire-enrich v2 Example - AI-powered data enrichment tool that transforms emails into rich datasets. See repo
• Enhanced Crawl Status & Warnings - Real-time status updates with clear feedback for robots.txt limitations and low-result scenarios
• 20+ Self-Host Improvements - Major stability and functionality upgrades for self-hosted deployments

Read the full changelog here

• Date: September 12, 2025

v2.2.0 Launch

New Features

• MCP version 3 is live. Stable support for cloud mcp with HTTP Transport and SSE modes. Compatible with v2 and v1 from.
• Webhooks: Now we support signatures + extract support + event failures
• Map is now 15x faster + supports more urls
• Search reliability improvements
• Usage is now tracked by API Key
• Support for additional locations (CA, CZ, IL, IN, IT, PL, and PT)
• Queue status endpoint
• Added maxPages parameter to v2 scrape API for pdf parsing

Read the full changelog here

• Date: August 29, 2025

v2.1.0 Launch

New Features

• Search Categories: Filter search results by specific categories using the categories parameter:
  • github: Search within GitHub repositories, code, issues, and documentation
  • research: Search academic and research websites (arXiv, Nature, IEEE, PubMed, etc.)
  • More coming soon
• Image Extraction: Added image extraction support to the v2 scrape endpoint.
• Data Attribute Scraping: Now supports extraction of data-* attributes.
• Hash-Based Routing: Crawl endpoints now handle hash-based routes.
• Improved Google Drive Scraping: Added ability to scrape TXT, PDF, and Sheets from Google Drive.
• PDF Enhancements: Extracts PDF titles and shows them in metadata.
• Map endpoint supports up to 100k results.

Read the full changelog here

• Date: August 19, 2025

Firecrawl v2

Key Improvements

• Faster by default: Requests are cached with maxAge defaulting to 2 days, and sensible defaults like blockAds, skipTlsVerification, and removeBase64Images are enabled.

• New summary format: You can now specify "summary" as a format to directly receive a concise summary of the page content.

• Updated JSON extraction: JSON extraction and change tracking now use an object format. The old "extract" format has been renamed to "json".

• Enhanced screenshot options: Use the object form.

• New search sources: Search across "news" and "images" in addition to web results by setting the sources parameter.

• Smart crawling with prompts: Pass a natural-language prompt to crawl and the system derives paths/limits automatically. Use the new crawl-params-preview endpoint to inspect the derived options before starting a job.

Read the full changelog here

We're excited to announce the release of Firecrawl v1.14.0, packed with cool updates.

v1.14.0 Launch

• Authenticated scraping (Join the waitlist here)
• Zero data retention for enterprise (Email us at help@firecrawl.com to enable it)
• Improved p75 speeds
• New MCP version w/ maxAge + better tool calling
• Open Researcher Example (Open Source Researcher). See repo
• And so much more. Check out here for all the details 🔥

We're excited to announce the release of Firecrawl v1.13.0, packed with awesome features.

v1.13.0 Launch

• Added AU, FR, DE to Enhanced Mode
• Crawl subdomains with allowSubdomains
• Google slides scraping
• Generate a PDF of the current page. See docs
• Higher res screenshots with quality param
• Weekly view for usage on the dashboard
• Fireplexity Example (Open Source Perplexity). See repo
• And more!