Back to vendors
B

Bardeen

Visit site
Entry priceFrom $10/mo · free tierFull pricing detail

Browser-native Chrome extension that scrapes lead data from any site, enriches and qualifies it with AI, and exports to Sheets, Airtable, Notion or CSV.

Bardeen is a browser-native automation tool that runs as a Chrome extension, which is its defining design choice: because automations execute inside the browser you are already using rather than on a remote server, Bardeen can act on any site you can see, including the many web apps that expose no public API. The trade-off is that this generally requires the browser to be open, though paid plans can run scheduled automations in the cloud when it is closed.

The product's centre of gravity is lead research for go-to-market teams. A point-and-click scraper pulls structured data off a page, a list of profiles, a company directory, a property listing, and a web search tool replaces manual lookups. Enrichment resolves validated emails and phone numbers against external data providers.

AI tools sit on top of the extracted data: an AI Qualifier scores each lead against criteria you define and shows its reasoning, and an AI text generator drafts personalised emails, content and social posts from the same data. Automations are composed in Bardeen Studio as Playbooks, triggered from the page you are on or from a browser context menu, and as Autobooks that run on a schedule or when a condition is met.

Finished data leaves as CSV or lands in Google Sheets, Airtable or Notion. Getting it into a CRM is a file import rather than a live sync: the vendor documents pushing to Salesforce, HubSpot, Pipedrive, Apollo and similar tools by exporting a CSV and using each destination's own import feature. A browsable template library, filterable by target website, covers common scraping and research jobs, and packaged solution sets are offered for solopreneurs, recruiting, real estate, private equity, music and event services.

Bardeen holds SOC 2 Type 2, CASA Tier 2 and Tier 3, and states that data from connected applications is persisted only in your local browser storage and never on its cloud servers, with cloud runs held stateless and erased once complete. Bardeen Inc. also sells a separate enterprise product, WIQ, on its own site at getwiq.ai, and that product's capabilities are graded on its own entry in this index rather than here. Bardeen itself is best understood as a browser-based research and tool for revenue teams rather than an autonomous agent platform, and it runs on credit-based pricing.

Vendor details

Canonical URL

https://www.bardeen.ai

Category

GTM / revenue agent

Funding status

Bardeen Inc., San Francisco. Backed by FirstMark and Insight Partners. The same legal entity also operates WIQ, a separate enterprise product at getwiq.ai.

Company status

independent

Use cases & customers

Target customers

knowledge workersSMB

Deployment options

SaaSbrowser extension

Integrations

Export to CSV, Google Sheets, Airtable and Notion. Connected apps via OAuth cover Gmail, Google Drive, Google Calendar, Google Sheets, Zoom and Slack. CRM and outreach tools including Salesforce, HubSpot, Pipedrive, Apollo, Outreach, Salesloft, Lemlist, Monday.com, SmartSuite, ClickUp and Zoho are reached by exporting a CSV and importing it using those tools' own import features, not by a native connection.

In practice

You spend hours copying LinkedIn profiles and company details into a spreadsheet or CRM. Bardeen's point-and-click scraper pulls the structured data off the page and drops it where you need it, no code required.

You want to automate a web app that has no API. Because Bardeen runs as a Chrome extension inside your own browser, it can act on any site you can see, clicking, filling, and moving data the way you would by hand.

Your prospecting means researching each lead before outreach. Bardeen's Browser Agents can visit a company's site, find and summarize details like its value proposition, qualify the lead against your criteria, and push it into your CRM.

Agentic Index coverage score

6.5 / 14 capabilities · 46%

Integrations & Tool Calling Partial

The headline integrations fail the threshold clause outright: they hand a result back to a person rather than letting the agent take authenticated action. The vendor's integrations page displays Salesforce, HubSpot, Pipedrive, Apollo, Outreach, Salesloft, Lemlist, Monday.com, SmartSuite, ClickUp and Zoho, but the copy says data is pushed to those tools or any platform that supports CSV, and every how-to link beside the logos points at the destination vendor's own CSV import documentation.

That is a file export the customer imports by hand. What remains is authenticated but short and fixed: export into Google Sheets, Airtable and Notion, and OAuth connections to Gmail, Google Drive, Google Calendar, Google Sheets, Zoom and Slack, with no custom connector or user-defined action documented. Downgraded from an April Full that carried no vendor evidence and claimed over a hundred apps.

Sourcebardeen.ai/integrations and bardeen.ai/securityread 2026-09-01

Workflow Orchestration Partial

Multi-step automation is real but shallow. Playbooks chain actions and are built in Bardeen Studio, where AI actions such as the Qualifier and the Text Generator are wired in alongside scraping and export steps, and Autobooks run the same chains on a schedule or trigger.

Held at Partial because the composition surface is thin against this axis: no branching or conditional logic, no error handling or retry behaviour, no sub-workflow composition and nothing multi-agent was documented on any page reached.

The pipeline the vendor markets is a fixed shape, scrape then qualify or generate then export, rather than an orchestration layer the customer designs freely.

Sourcebardeen.ai/ai, bardeen.ai/integrations and bardeen.ai/securityread 2026-09-01

Knowledge Grounding & RAG Unable to verify

Grounding runs on the public web rather than on the customer's own knowledge. Web Scraper extracts from live pages, Web Search replaces manual lookups, and Enrichment resolves contact details against external providers. Import from Google Sheets, Notion and Airtable is data movement into a scraping run, not a maintained retrieval structure the agent grounds on.

No knowledge base, no document or file ingestion, no vector store and no persistent corpus of customer material was reached across the product pages, integrations catalogue or security page. Qualification criteria the customer types into the AI Qualifier are per-run instructions rather than a grounding source. Confidence normalised from a non-canonical value; the grade is unchanged.

Sourcebardeen.ai/ai, bardeen.ai/scraper and bardeen.ai/integrationsread 2026-09-01

Human Oversight & Guardrails Unable to verify

No shipped approval or guardrail mechanism is documented. Oversight here is structural rather than gated: automations are invoked by the user from the page they are on and most run in the user's own open browser, so a person is present by architecture. That is the operating mode, not a mechanism the vendor ships. No per-action approval step, no run-confirmation gate, no permission scoping over what an automation may touch and no configurable guardrail was reached across the product pages, integrations catalogue, security page or site FAQ.

Notably the same company's separate product does advertise pause-for-review checkpoints and escalation paths as differentiators, which cannot be credited to this record. Graded from a pass that reached first-party material and returned empty on this axis rather than from a failed pass; the help centre at support.bardeen.ai was not opened and would settle it. Downgraded from an April Full that carried no vendor evidence.

Sourcebardeen.ai, bardeen.ai/integrations and bardeen.ai/securityread 2026-09-01

Security, Identity & Governance Partial

Attestation is strong and the control surface is absent, which is the opposite failure to most records at this grade. Certified: AICPA SOC 2 Type 2 since 1 April 2024, CASA Tier 2 and Tier 3 against the App Defense Alliance assessment built on OWASP ASVS, and GDPR adherence, all badged on the site.

Engineering practice is documented in unusual detail: TLS 1.2 in transit, 256-bit AES at rest, infrastructure as code with versioned changes, peer review on all code changes, automated vulnerability scanning, and a published security contact for vulnerability disclosure.

What is not documented anywhere on the security page, the FAQ or the product pages is a customer-facing identity or governance control: no SSO, no SAML, no role-based access, no audit log and no permission model, on a product the pricing describes as including unlimited users.

One defect on the governance surface itself: the security page links its privacy policy and contact address to a different domain, getwiq.ai, while the site footer links to bardeen.ai/privacy-policy, so the two routes a buyer would take to the same policy disagree. Partial because the convention requires attestation plus at least one named customer-facing control for Full. Downgraded from an April Full that carried no vendor evidence.

Sourcebardeen.ai/securityread 2026-09-01

Observability & Auditability Partial

The AI Qualifier is documented as showing its work when it scores a lead against the customer's criteria, which is per-run explainability over the one decision the product makes autonomously, and that is what holds this cell above the floor.

Held at Partial because nothing wider is documented: no run history, no execution log across Playbooks and Autobooks, no audit trail over who ran what, and no failure or error surface the customer inspects was reached on either pass.

The extensive logging the security page describes is vendor-side telemetry through Amplitude, Statsig and Sentry, collected explicitly without user data, and is not a customer surface.

Sourcebardeen.ai/ai and bardeen.ai/securityread 2026-09-01

Memory & State Persistence Partial

State persists on the customer's own machine: scraped and connected-app data is held in local browser storage across sessions, with the extension declaring the unlimitedStorage permission because the default cap is too small for the data volumes involved, and browser history and bookmarks are read locally to suggest relevant automations. Held at Partial because the persistence is a cache and a configuration store rather than agent memory.

Playbook and Autobook definitions held server-side are customer-authored configuration, which this axis excludes, and the vendor states explicitly that a cloud instance running an automation with the browser closed has no storage capability and erases all data once the run finishes. No accumulation across runs and no memory the agent writes to and reads back is documented.

Sourcebardeen.ai/securityread 2026-09-01

Deployment & Data Residency Full

Execution happens on the customer's own machine and the vendor documents the data path in detail. App data from connected applications is persisted only in local browser storage and never on vendor cloud servers, with the exchange running directly between the customer's browser and the third-party application.

Where automations are set to run with the browser closed, a cloud instance is created per trigger and the vendor states it has no storage capability, so all data is erased once the run completes. Only connected-app configurations and playbook definitions are held server-side. Graded Full on customer-environment execution, the browser-extension form of on-premises. The limitation, stated plainly: there is no region selection for the metadata that is stored, no VPC option and no self-hosted backend.

Sourcebardeen.ai/securityread 2026-09-01

Prebuilt Agents, Templates & Packs Full

A browsable template catalogue is published at bardeen.ai/templates, carries a top-level slot in the site navigation, and is filterable by target website, with GitHub, YouTube, Crunchbase, Monster, Upwork, Product Hunt, Realtor, Facebook, Instagram and Zillow surfaced directly in the footer and a show-all view behind them.

These are packaged scrapers and automations a customer adopts and runs rather than configurations the vendor assembles for them. Six packaged vertical solution sets sit alongside the catalogue for solopreneurs, music, real estate, recruiting, private equity and event services. This is the browsable catalogue the Pack bar treats as evidence, and it is the thing the two GTM records reviewed before this one lacked.

Sourcebardeen.ai/templatesread 2026-09-01

Triggers & Channel Coverage Full

The agent starts without a person prompting it. Autobooks run on a schedule or when a condition is met, with the vendor giving an inbound email as the worked example, and on paid plans those runs continue with the browser closed by spawning a cloud instance each time the condition fires. Both a schedule and an inbox rule are named trigger types.

On the channel side the agent meets the user in the browser where the work is already happening, overlaying its interface on the page in view, with browser notifications for time-sensitive prompts and a custom context-menu entry, and it reaches into Gmail, Google Calendar and Slack through OAuth. Manual invocation from the page sits alongside those rather than replacing them. The published Partial marker, one surface with no event triggers and no way into email or a team channel, does not apply here.

Sourcebardeen.ai/security and bardeen.ai/integrationsread 2026-09-01

Model Flexibility & Routing Unable to verify

The AI Tools page states that large language models are used to categorise leads and compose outreach, which confirms models are in the product but names none of them. No provider is disclosed anywhere on the product pages, the integrations catalogue or the security page, and no model selector, per-automation model choice, bring-your-own-key or routing control is offered to the customer.

The security page enumerates the third-party services the vendor uses for telemetry, A/B testing and crash reporting but lists no model provider, so even the sub-processor route gives nothing. Not documented as first-class, with the absence inferred from silence rather than from a stated single provider. Confidence normalised from a non-canonical value; the grade is unchanged.

Sourcebardeen.ai/ai and bardeen.ai/securityread 2026-09-01

APIs, SDKs & MCP Extensibility Unable to verify

No API, SDK, MCP server or webhook is documented. The product pages, the integrations catalogue, the security page and the site FAQ were all read: the only routes in or out are the Chrome extension itself, OAuth connections to a short list of apps, and CSV export. The extension is a distribution surface rather than an interface the customer's own systems can call, and CSV is a file format.

Worth recording that the same company's separate product does ship an MCP server and advertises being callable from Claude, Workato and Copilot, which belongs to that product and not to this record. Graded from a pass that reached first-party material and returned empty on this axis rather than from a failed pass; the help centre at support.bardeen.ai was not opened. Downgraded from an April Partial that carried no vendor evidence.

Sourcebardeen.ai/integrations, bardeen.ai/ai and bardeen.ai/securityread 2026-09-01

Testing, Debugging & Optimization Unable to verify

Nothing the customer can test, debug or optimise is documented. No evaluation surface, scored run, comparison between two versions of a scraper or automation, dry-run or preview mode before an Autobook goes live, and no accuracy measurement over extracted or qualified data was reached across the product pages, integrations catalogue, security page or FAQ.

The nearest thing is the AI Qualifier stating that it shows its work, which is per-run explainability credited on Observability rather than a test the customer runs, and it is spent there. Graded from a pass that reached first-party material and returned empty on this axis rather than from a failed pass; the help centre at support.bardeen.ai was not opened. Downgraded from an April Partial that carried no vendor evidence.

Sourcebardeen.ai/ai and bardeen.ai/scraperread 2026-09-01

Browser & Computer Use Full

The clearest positive case for this axis in the lane.

The product is a Chrome extension that operates sites the vendor does not control and that expose no programmatic interface, and the vendor documents exactly how: host permission to lay its own interface over whatever page the user is browsing, activeTab and tabs permissions to open, close and switch browser tabs and to run background scraping, capture page snapshots, and webNavigation to drive OAuth flows.

Point-and-click selection extracts structured data from any page, and the vendor positions this as its defining design choice, since running inside the browser lets it act on web apps that publish no API. Sourcing runs against live sites rather than a static database.

Sourcebardeen.ai/security and bardeen.ai/scraperread 2026-09-01

The Agentic Index coverage score grades every vendor Full, Partial or Unable to verify against the same 14 buyer facing capabilities, from public evidence only. Each capability links to how all vendors in the index score on it. How this evidence is graded

Recent platform changes

2026-08-27·Pricing / packagingVerified

Bardeen introduced a revamped pricing model that eliminates per-seat licensing, allowing unlimited team members on all plans. The update also rolls out new Business and Enterprise tiers designed to provide more predictable costs and flexibility.

Bears on: Pricing / packaging

View source
2026-08-27·Funding / partnershipVerified

Bardeen announced a $3 million strategic investment from Dropbox and HubSpot alongside the launch of its Business-Ready AI agent. The release evolves the platform from standard workflow automation into an instructable agent built for generative AI use cases.

Bears on: Agent capability

View source
View all 2 changes for Bardeen →Tracked since Aug 2026 · Verified from public vendor sources

Pricing

From $10/mo · free tier

credits

Free tierTrial available

Included quota

Free: 100 credits/mo (always), non-premium actions, basic scraping/extraction, manual triggers, new builder, unlimited team members - no scheduled automations, premium integrations, or AI actions. Basic/Pro ($10/mo): premium integrations (Salesforce/HubSpot), scheduled + cloud automations, AI-powered actions/enrichment, larger credit pool. Premium ($50/mo, $480/yr): ~1,000 credits/mo, full automation. Business/Enterprise (custom): shared playbooks, admin controls, team analytics, SSO, dedicated GTM consultant, solutions engineering, priority support. 1,000+ pre-built Playbooks (templates).

What is public

Bardeen (bardeen.ai - browser-based (Chrome-extension) automation + AI-agent platform for scraping, workflow automation, and GTM/prospecting) is credit-metered with unlimited seats on every plan. Free: 100 credits/mo (always free, non-premium actions). Basic/Pro: $10/mo (premium integrations, scheduled/cloud automations, AI actions). Premium: $50/mo (1,000 credits, $480/yr ~ $40/mo). Enterprise: custom. Under 'Pricing 2.0' (announced effective Dec 1, 2024), every action now costs a flat 1 credit (enrichment rows 3 credits); imports/utilities/CSV exports are free. Some 2026 sources show alternate tier framing (Pro $10/Business $15/user, or Starter $99/Teams $500/Enterprise $1,500).

Billing mechanics

Credit-metered subscription with UNLIMITED seats on all plans (since 'Pricing 2.0', effective Dec 1, 2024). The simplified model: every action that creates a row = 1 credit; enrichment rows = 3 credits; importing data, utilities, and CSV downloads are FREE. You always get 100 free credits per period; unused credits expire at the end of each billing period (no rollover). Paid plans add larger credit pools; overage requires buying additional credit packs (~$0.02/credit, no auto-overflow). Annual billing discounts the monthly rate.

Cost watchouts

Credit MATH is the real cost, not the headline price - enrichment triples credit burn (3 credits/row) and high-frequency cloud automations (running 500+ times/day) deplete credits fast (one tested 5-person team's true cost rose ~$10 to $25/user/mo after enrichment overages); credits don't roll over (expire each period); CHROME-ONLY (no Firefox/Safari/Edge - limits team adoption); browser-scraping playbooks are FRAGILE - LinkedIn/Salesforce UI changes break them, requiring repair every few weeks; Bardeen has NO own B2B data (it scrapes visible pages)

Variable cost rationale

Usage-driven via credits - cost tracks your monthly ROW volume (1 credit/action-row, 3/enrichment-row), not the sticker price; cloud/scheduled automations and enrichment are the big burners, and with no rollover + paid overage packs, spiky usage inflates the effective bill well above the base plan

Additional watchouts

Chrome-only (excludes Firefox/Safari/Edge users); browser-scraping playbooks break on target-site UI changes (LinkedIn especially - expect periodic rebuilds, no auto-monitoring); enrichment + cloud automations burn credits fast (budget the credit math before scaling); no native B2B data (scrapes what's visible); credits expire each period (no rollover); for production signal-driven outbound, dedicated GTM platforms avoid the maintenance overhead

Overage / add-ons

100 free credits each period; unused credits EXPIRE at period end (no rollover). Exceeding your plan's credits requires purchasing additional credit packs (~$0.02/credit) - there's NO automatic overflow or auto-upgrade, so cloud/high-frequency automations can deplete credits mid-cycle and pause until you top up. Standard action = 1 credit; enrichment = 3 credits; imports/utilities/CSV = free.

Sales call required

Mixed (some tiers require a call)

Free / trial

Free; 7-day trial

Lowest paid plan

$10/mo

Commercial notes

Founded 2020 by Artem Harutyunyan & Pascal Weinberger; ~$25.3M raised (incl. Dropbox Ventures + HubSpot Ventures); repositioned May 2025 to a 'Work Intelligence Platform' / 'the AI agent to automate automations,' with center of gravity now on sales/GTM prospecting; strongest at browser scraping + data extraction + enrichment; 1,000+ pre-built Playbooks; competes with Zapier ($29.99, cloud/8,000+ apps), Clay ($149), PhantomBuster ($56), Apify ($49) - cheaper but narrower (browser-bound)

Key ambiguities

Tier framing/credit allotments are cited inconsistently across 2026 sources - some show Pro $10 (500 credits)/Business $15/user; others Basic $10 (100 credits)/Premium $50 (1,000 credits, $480/yr); a third set shows Starter $99 (15,000 credits/yr)/Teams ~$500/Enterprise ~$1,500; annual-discount % varies (17%-33% cited); the official page confirms the single-credit model + 100 free credits but the exact paid-tier names/credit counts drift between sources

Cancellation / refund

Free 100 credits/mo always (no card for the free tier); 7-day Pro trial (self-serve) and a 'request' 14-day team trial via sales; paid plans self-serve monthly or annual (annual cheaper); buy additional credit packs as needed; Business/Enterprise custom-contracted; unlimited seats on every plan

Support SLA / resale

Community/standard support (Free/Pro); Business/Enterprise add priority support, admin controls, team analytics, shared playbooks, solutions engineering, and a dedicated GTM consultant; SSO on Enterprise; SOC 2 Type II, GDPR, and CASA Tier 2/3 certified; integrates with Gmail/Slack/Google Sheets/Salesforce/HubSpot; runs as a Chrome extension (local processing option)

Missing data

Paid-tier names/credit counts and annual-discount % are cited inconsistently across sources (Pro/Premium vs Basic/Business vs Starter/Teams framings; 17%-33% discount). Seed 'Basic $10/mo (100 credits); Premium $50/mo (1,000 credits, $480/yr); Enterprise custom' aligns with the consumer-tier framing - note 'Pricing 2.0' (announced effective Dec 1, 2024) simplified to single-credit pricing (1 credit/action, 3/enrichment) with unlimited seats on all plans, and 100 free credits always.

Agentic Index verified 2026-06-25

Alternatives to Bardeen

The closest documented capability profiles to Bardeen among GTM and revenue agents tracked by Agentic Index, ordered by similarity on the same 14 point evidence the rankings use. No vendor pays for placement.

  • AnyBiz4.0 / 14Adds documented Knowledge Grounding & RAG
  • Floworks5.0 / 14Adds documented Knowledge Grounding & RAG
  • Laxis7.0 / 14Adds documented Knowledge Grounding & RAG and Human Oversight & Guardrails, among others
  • Nooks7.0 / 14Adds documented Knowledge Grounding & RAG and Human Oversight & Guardrails, among others
  • AiSDR7.5 / 14Adds documented Knowledge Grounding & RAG and Human Oversight & Guardrails, among others
  • BambooBox5.5 / 14Adds documented Knowledge Grounding & RAG

Similarity is computed from each vendor's Agentic Index coverage score evidence, axis by axis, not from the totals. How this evidence is graded

Contact us

Found a vendor we missed? Have feedback on the index? We'd love to hear from you.