Releases Index

Browser Automation & Scraping

Headless browsers, crawlers, and scraping APIs — the data layer for AI agents.

This week's digestNov 9 – 15, 2026

Agents get richer tools while browser sessions get organized

Firecrawl pushed its agent surface hard this week, expanding its company and AI/ML research data into ChatGPT and Claude and shipping a rebuilt MCP with thousands of tools, while Browserbase added session labels and dashboard-managed secrets for keeping automation runs tidy and secure.

Read the digest6 releases covered
Week of Nov 96 releases
DigestAgents get richer tools while browser sessions get organizedRead →
ThuNov 12, 20261 release
Firecrawl
Daily summary

Firecrawl's Alexandria expands to 75M company profiles

Firecrawl Alexandria now returns firmographic data — headcount, revenue, funding, and tech stack — for more than 75 million companies in about half a second, backed by 3,000+ tools and provider waterfalls.

  • Firecrawl Alexandria now resolves firmographic lookups for over 75 million companies, returning headcount, revenue, funding, and tech stack in roughly half a second.
WedNov 11, 20262 releases
BrowserbaseFirecrawl
Daily summary

Firecrawl MCP v3 brings 3,000+ tools and far higher rate limits

Firecrawl's MCP server expanded to over 3,000 tools with automatic provider failover, 60x higher rate limits, 4x more usage, and parallel subtasks across every harness.

  • Firecrawl MCP v3 gives agents access to 3,000+ tools, falling back automatically across providers when one fails.
  • The release raises rate limits 60x and usage allowance 4x, and adds parallel subtasks, available in every harness.

Sessions get labels for filtering and organization

Session labels let you organize browser sessions by workflow, environment, website, or automation version and filter for them in the dashboard. Labels can be added via the API or in the dashboard, are shared within a project, and can be renamed or removed.

Details
Session labels
TueNov 10, 20261 release
Firecrawl
Daily summary

Firecrawl adds an AI/ML research index for agents

Firecrawl launched an AI/ML research index that lets ChatGPT, Claude, or any agent harness with Firecrawl find the right paper, including same-day postings, and read its full text.

  • Firecrawl's AI/ML research index lets agents find the right paper and read its full text, including papers posted the same day.
MonNov 9, 20262 releases
BrowserbaseFirecrawl
Daily summary

Firecrawl adds resume-to-role matching and exec contact lookup

Firecrawl shipped job search tools that match a resume against more than 7 million open roles and surface executives to contact from a database of over 250 million people, with context on their interests for outreach.

  • Firecrawl now matches a resume against 7M+ open roles and finds executives to email from a 250M+ person database, showing what they care about before outreach.

Job search tools in ChatGPT and Claude

New job search tools match a resume against more than 7 million open roles and help find executives to email from a database of over 250 million people, showing what they care about before outreach.

Details
Job search tools for ChatGPT & Claude
Week of Nov 28 releases
Weekly digest
Agent data gets deeper while agent reach gets governed

Firecrawl expanded its Alexandria corpus to SEC filings, life sciences research, and election records—each readable directly inside ChatGPT and Claude—while Browserbase introduced Navigation Policies to control which sites browser agents can reach and Apify hardened its SDK against restarts.

Read the digest →
SunNov 8, 20261 release
Firecrawl
Daily summary

Firecrawl celebrates 200K GitHub stars with plushies

Firecrawl marked 200K GitHub stars by producing 2,000 plushies, claimable through the Firecrawl plugin while supplies last, in a day with no shipping changes to the product itself.

  • Firecrawl reached 200K GitHub stars and is giving away 2,000 plushies, claimable via the Firecrawl plugin while supplies last
ThuNov 5, 20262 releases
FirecrawlApify
Daily summary

Apify SDK persists event charges across restarts

The Apify SDK now keeps charged event counts across container restarts and emits an exit event reliably, along with fixes for local input reads under non-filesystem storage clients and request-list URL matching.

  • Apify SDK now persists charged event counts across container restarts, so Actors no longer lose event charges when a container is recycled
  • Actors emit Event.EXIT on exit, and local Actor input can be read from the storage directory when using non-filesystem storage clients
Apify SDKv4.2.1
WedNov 4, 20262 releases
BrowserbaseFirecrawl
Daily summary

Firecrawl brings 41M life sciences papers into ChatGPT and Claude answers

Firecrawl made 41 million life sciences research papers queryable in a single step, so asking ChatGPT or Claude about a drug, disease, or gene now returns answers drawn directly from that literature.

  • Firecrawl's life sciences index of 41M+ research papers is now reachable through one-step ChatGPT and Claude queries on drugs, diseases, and genes

Browsers persist across disconnects; sessions survive SDK crashes

browser.disconnect() now closes Stagehand's connection without closing the browser or session, and you can reattach later via localBrowser.connect({ cdpUrl }) or browserbase.connect({ sessionId }). The Stagehand extension also cleans up when the SDK process crashes on Chrome 137+, and keepAlive now works on local browsers.

Details
Stagehand 4.4
TueNov 3, 20262 releases
BrowserbaseFirecrawl
Daily summary

Firecrawl benchmarks search APIs on legal queries

Firecrawl published a retrieval benchmark scoring three search APIs against 581 legal research requests judged against the official source documents, with its Government & Legal Index leading on both gold@1 and gold@10 accuracy.

  • Firecrawl released a legal-retrieval benchmark that grades three search APIs on 581 legal research requests by whether results surface the authoritative document, with its Government & Legal Index ahead on gold@1 and gold@10.

Navigation Policies restrict where browser agents can go

Navigation Policies let you allow or block categories of sites, add rules for specific domains, and reuse one policy across sessions. A policy is created through the API and its ID passed as browserSettings.riskPolicyId when starting a session; changes apply to new sessions only and the feature requires support to enable per project.

Details
Navigation Policies
MonNov 2, 20261 release
Firecrawl
Daily summary

Firecrawl adds free midterm election records for agents

Firecrawl published a free Election Records dataset in Alexandria covering all 470 House and Senate races, letting AI agents query live results as states report, campaign funding, and incumbent voting records.

  • Firecrawl's free Election Records dataset lets AI agents query all 470 House and Senate races, including results as each state reports, campaign funding, and incumbent voting records.
Week of Oct 266 releases
SunNov 1, 20262 releases
Firecrawl
Daily summary

Firecrawl raises Government Index limits and brings enrichment flows to chat assistants

Firecrawl tripled Government Index search throughput to 300 requests a minute on every plan, still credit-free, and launched headless Clay so agents in ChatGPT and Claude can build and run people and company enrichment flows with 1,000 free credits a month.

  • Firecrawl now allows 300 Government Index searches a minute on all plans and continues to charge no credits for them.
  • Headless Clay lets agents inside ChatGPT and Claude build and run enrichment flows over 250M+ people and 26M+ companies, with 1,000 free credits a month.

Firecrawl Changelog

2 updates

Headless Clay lets agents build and run enrichment flows in ChatGPT and Claude, searching over 250 million people and 26 million companies with 1,000 free credits a month.

Details
Headless Clay

Government Index searches now allow 300 requests a minute on every plan, three times the previous limit. The searches still consume no credits.

Details
Government Index
FriOct 30, 20261 release
Apify
Daily summary

Apify SDK shifts actors to spend-based task limits

The Apify SDK deprecates max_items in favor of a max_total_charge_usd spend cap on start_task and call_task, and makes named child runs survive migrations and graceful aborts.

  • Apify SDK now lets Actors cap task costs with max_total_charge_usd, with max_items deprecated in its favor
  • Named child runs are exposed via Actor.child_runs, reattach after migration or resurrection, and abort cleanly with the parent
  • Apify SDK adds support for Python 3.15
Apify SDKv4.2.0
ThuOct 29, 20261 release
Firecrawl
Daily summary

Firecrawl brings web access to Vercel's eve agent framework

Firecrawl shipped an integration for eve, Vercel's open-source agent framework, letting agents gain web access through Alexandria with a single command.

  • Firecrawl is now an integration in eve, Vercel's open-source agent framework, adding web access to agents via Alexandria with one command.
WedOct 28, 20261 release
Firecrawl
Daily summary

Firecrawl feeds 400,000+ vulnerabilities to coding agents

Firecrawl added cybersecurity data to its Alexandria index, letting Codex and Claude Code agents check a codebase's dependencies against more than 400,000 known vulnerabilities and patch them before attackers can exploit them.

  • Firecrawl's Alexandria now includes cybersecurity data, so Codex and Claude Code agents can scan project dependencies against 400,000+ vulnerabilities and patch them.
TueOct 27, 20261 release
Firecrawl
Daily summary

Firecrawl brings 250M-profile recruiting data to ChatGPT and Claude

Firecrawl launched Directory, which lets users describe a role in ChatGPT or Claude and get a candidate shortlist drawn from 250M+ profiles in Alexandria, including experience and work emails, at no cost.

  • Firecrawl's Directory gives ChatGPT and Claude users free recruiting shortlists from 250M+ Alexandria profiles with experience and work emails