Lindy Review 2026: The Agent Builder That Became a Slack-Native AI Teammate and Now Bills by the Seat and the Credit
In 2025 Lindy was "build your own AI employees": Lindy 3.0 launched in August 2025 as a prompt-to-agent builder. On Oct 3, 2026 lindy.ai sells "the AI teammate that will 3x your output" that lives in Slack, iMessage and email, with a seat for every active user from $29.99 a month, a pooled credit balance that does not roll over, and a free tier of $50 in credits that lasts seven days. This review checks what is true today against Lindy's own pricing page, docs, security page and changelog, adds what real users say on Trustpilot, Product Hunt and Hacker News, and is blunt about where credits, cancellation and thin documentation of the older agent builder cost trust.
| Ownership | Independent private company; no acquisition found, funding not verified |
| Free tier | $50 credits for 7 days, no card; not a standing free plan |
| Cheapest paid plan | $29.99/mo per user, 3,000 credits |
| Security | SOC 2 Type II (vendor); HIPAA BAA on Enterprise only |
| Trustpilot record | 1.7/5 as displayed, 40 reviews |
Lindy stopped being the tool you build agents in and became a teammate you hire into Slack. That is a different purchase, and it bills like one.
"If your team lives in Slack and wants one assistant that reads the channels, joins the meetings and drafts the emails, Lindy now sells exactly that. If you came for a drag-and-drop agent builder you can predict, the product on the front page is no longer that, and the credit meter is where most of the complaints live."
My honest read: Lindy has three jobs today. It is a shared Slack teammate (@mention it in a thread and it pulls from your approved sources, then writes the answer, the deck or the dashboard). It is a personal assistant with a daily brief, an inbox that labels and drafts in your voice, and meeting notes (by iMessage, SMS, a Gmail Chrome extension or the web app). And it is a routine runner that does recurring work on a schedule or trigger. The first two are the story now. What changed is the commercial model: a seat for every active user, a pool of credits that does not roll over, and a free tier that is $50 of credits for seven days rather than a standing plan. This is a VERIFIED EDITORIAL REVIEW: Lindy's pricing page, docs, security page, changelog and sitemap were checked on Oct 3, 2026, alongside Trustpilot, Product Hunt and Hacker News sentiment. I did not run a controlled hands-on test.
One workspace, two Lindys: a shared one in your Slack channels and a personal one in your DMs, texts and inbox.
The docs open with one sentence: "Lindy is an AI employee. It lives in your Slack, connects to your tools, and comes back with the work done." Underneath that sit four workspace assets the docs keep returning to. Skills are named playbooks Lindy picks up on its own. Routines are standing work on a schedule or trigger. Memory is a versioned file system you can open and edit. Meetings is a recorded, searchable library. The docs are explicit that both Lindys draw from the same credit pool: the shared one answers your team in channels, and the private one runs each person's brief, alerts and recaps.
Ownership: I found no acquisition and I did not verify any funding figure, so none is cited. Lindy's own site names customers such as Shopify, Apple, Adobe, McKinsey, NVIDIA, Harvard, Airbnb, Autodesk and Stanford, which I treat as marketing, not audited. If you want the neighbouring, deterministic route, JAVIS's Zapier review and Make review cover the classic trigger-and-action builders.
A well-known name in the AI-agent space whose public footprint is mostly its own claims.
Read Oct 3, 2026, 1.4K followers. A very small, self-selected sample, not a market-share measure.
78% of reviews are one star. Reviews are not users.
Shown on lindy.ai. G2 was not available to us, so I could not confirm the number or the review count. Treat as unverified.
The homepage and docs say 1,000+; the pricing page says 1,500+. Same company, two counts. Neither is audited.
| Date | Event | Source |
|---|---|---|
| 2023-03-22 | "Meet Lindy": launched as an AI personal assistant | Lindy blog, via Hacker News item 35267091 |
| 2024-02-15 | "Show HN: Lindy, build your own AI employees with no-code" (3 points) | Hacker News item 39385343 |
| 2025-07-22 | 500+ new actions across 200+ apps through a Pipedream partnership | lindy.ai/changelog |
| 2025-08-04 | Lindy 3.0: Agent Builder ("vibe coding for agents"), Autopilot cloud computer, Team Accounts | lindy.ai/changelog |
| 2025-08-28 | Lindy Build, an app builder, announced | lindy.ai/changelog |
| 2025-10-21 | Last entry on the public changelog page (retries, task search, version renaming) | lindy.ai/changelog |
| 2026-02 | Lindy Assistant launched on Product Hunt (ranked #4 per the page) | producthunt.com/products/lindy |
| 2026-09 | Docs describe Lindy Teammate: Slack-native, per-user Plus / Pro / Max seats, pooled credits | docs.lindy.ai/pricing (sitemap lastmod 2026-09-13) |
The honest summary of market position: Lindy has a recognisable brand and a sizeable marketing footprint, but I could not find audited usage, revenue or customer counts, and the independent review volume (7 on Product Hunt, 40 on Trustpilot) is thin. Nothing here ranks Lindy against Zapier, Make or n8n, whose scale numbers live in their own JAVIS reviews: Zapier, Make and n8n.
Fans like the finished work. The recurring complaints are credit burn, cancellation and support.
Reviewers quote "No working cancel option -- kept getting charged" (Jun 2026), "I managed to kill my months credits (3000) in one day!" (Aug 2026), "1000-4000 average credits per day" against an expected 600-800 (Aug 2026), a Gmail label bug that created a blank "undefined" label (May 2026), and support that "doesn't respond" (May 2026). Billing complainers are over-represented on review sites, so weigh the pattern more than the score.
Praise centres on time saved ("saves us hours every week" through pre-built workflows), the fact that it "performs tasks across different tools" rather than only generating text, a no-code interface, and a helpful Slack community. Minor friction: finding time to explore more automations, and occasional inaccuracies when AI rebuilds workflows. Note these reviews mostly describe the earlier agent-builder era.
One commenter (Jan 2025) recommended Lindy over a beta chatbot task feature, saying it sends "periodic email updates on specific topics and it works flawlessly" (no affiliation stated). Another (Feb 2025) wrote "We use lindy.ai" and expected it to be widely used. Volume is tiny: the three Hacker News stories I found that name Lindy in the title or URL scored 1, 5 and 3 points.
Free for a week, then a seat per person: $29.99, $99.99 or $199.99 a month, each with its own credit pool.
Free
$50 in credits, no credit card, set up in 2 minutes.
- Credits last 7 days only
- Pricing page: "Everything you need to get started"
- No standing free plan I could find
Plus
3,000 credits per seat per month, pooled across the workspace.
- 2 connected inboxes
- All models, model selection per task
- Shared credit pool, spend caps per teammate
- No computer use per the docs
Pro
15,000 credits per seat, "5x Plus" usage.
- 3 connected inboxes
- Computer use
- Live onboarding on Pro and up
- Same pooled model
Max
35,000 credits per seat, about 11x Plus per the docs.
- 5 connected inboxes
- Computer use
- Largest self-serve pool
- Top-ups at $10 per 1,000 credits
| Tier (pricing page) | Price | Credits | JAVIS arithmetic: price per 1,000 credits | Who it suits |
|---|---|---|---|---|
| Team entry (docs: Plus) | $29.99 | 3,000 | $10.00 | Trying it |
| Team 15,000 (docs: Pro) | $99.99 | 15,000 | $6.67 | Daily delegation |
| Team 35,000 (docs: Max) | $199.99 | 35,000 | $5.71 | Heaviest self-serve use |
| Top-up (any plan) | $10 | 1,000 | $10.00 | Topping up the pool |
| Overage if enabled (docs) | 2x regular credit cost | n/a | about $20.00 if a credit is about one cent | Keeping work running past the limit |
| Enterprise | Custom | Custom | Custom | SSO/SCIM, audit logs, HIPAA with a signed BAA, dedicated support |
Two details change the maths. First, "every active user in your workspace is billed": anyone who signs up, joins, or even @mentions Lindy in Slack takes a seat, with a seven-day free week for people who join through Slack and immediate billing for direct sign-ups. Second, monthly credits do not roll over, while purchased top-ups do. Check live pricing
Decide by where your team already talks and how predictable your workflows must be.
Lindy is built for exactly that. Start with the $50 free week in one channel.
That is the Zapier, Make or n8n shape. See the Zapier, Make and n8n reviews.
Credits are metered and work varies from 2 to 2,500 credits per task by Lindy's own table. Set spend caps and watch the first week closely, or choose a flat-priced tool.
Those sit on Enterprise only. The security page says controls are "mapped to HIPAA requirements", while a signed BAA is Enterprise.
Plus on its own may do it, but compare against the assistants in your existing email and calendar tools first.
Lindy's own Aug 31, 2026 blog still lists the Agent Builder and the app-builder page is live, but the pricing page does not say how agents are billed. Ask support before you pay.
You @mention Lindy in a thread; it searches your approved sources, calls your tools and posts finished work.
The docs describe the shared Lindy as the place "where most of a team's credits go, because everyone @mentions the same Lindy." It answers from sources an admin approves (channels, Drive folders), can build live pages called Artifacts (dashboards, reports, prototypes, each on its own link), and keeps going in a thread. Anything with outside impact waits for approval: sending an email, updating a ticket, posting to another channel, publishing a doc. Read-only lookups from approved sources do not need approval. The homepage demos show the intended flavour: "why did ad spend double", "what is blowing up the support inbox", "reconcile Stripe against the ledger". Those are vendor demos with sample data, not test results.
Lindy's own announcement of the product this review is about, the clearest primary-source picture of the shared-teammate idea.
Published by the official "Lindy" YouTube channel. Vendor announcement, so it shows the best case. I could not confirm its publish date from the channel listing.
Ask in plain words, let Lindy read and act, approve anything that writes, then check the credit bill.
Three habits matter. First, the usage warning: when a task starts drawing "significantly more resources than usual" Lindy stops and asks you to Continue or Stop in the same thread. Second, dry runs for routines: you can ask Lindy to "dry-run my morning briefing" and it shows what it would have done without sending or writing anything. Third, scoping team sources tightly, which the docs call "the rare lever that improves quality and cost at once".
Routines, a meeting recorder and a Gmail extension are the "do it while I am away" half.
A routine has three parts: a trigger (a schedule, or an event such as an email received, a Slack message, a calendar event about to start, a meeting ending), a prompt in plain English and a destination. Built-ins include daily briefs, email drafting, email labeling, urgent-email alerting, follow-up bumps, meeting recording, prep and scheduling. Personal routines can be converted to a Python script that runs instead of waking Lindy each time, which is "faster, cheaper, predictable", but scripts do not carry your connected-account credentials, so the docs say "Summarize my unread Gmail" would not work that way. The meeting recorder sends a notetaker to Google Meet, Zoom or Teams that joins visibly as a named participant; the docs still say to tell participants they are being recorded as local laws require.
Computer use (the agent operating a browser or computer for tools with no integration) is listed for Pro and Max in the docs. The pricing page also lists "Computer use" under the Team plan's included features without naming a tier, so I treat the docs table as the tighter statement and flag the mismatch. The Chrome extension is Gmail only today, with Outlook "coming soon". The homepage says "Computer use capability (Pro and up)", which matches the docs. Skills are labelled "Early access" and "rolling out gradually behind feature flags".
Lindy's own walkthrough of the personal-assistant side (brief, inbox, meetings), which is what the iMessage, email and routine features add on top of the shared Slack teammate.
Published by the official "Lindy" YouTube channel. Vendor demo; it shows the best case and pre-dates the current Teammate pricing, so check prices on the page, not in the video.
1,000+ integrations, any MCP server over streamable HTTP, and a browser for everything else.
Lindy says it connects to "1,000+ tools" (Gmail, Slack, Notion, HubSpot, Google Drive, GitHub) and supports MCP, so "any MCP server your team runs plugs straight in". The docs add two operational facts. You connect through OAuth or an API key, and every action for the tool "turns on at once", so there is no action-picking step. And MCP servers must expose a public HTTPS streamable-HTTP endpoint; on connect Lindy reads the catalog and creates an action for every supported tool, which "a large server" can make "well over a hundred". Personal connections are private; an owner or admin can share one with the whole workspace, after which "everyone's Lindy reads that tool as that account". If MCP is new to you, JAVIS's Claude MCP and Connectors guide explains the idea.
| Integration path | What it does | Status as read |
|---|---|---|
| Built-in integrations | Gmail, Slack, Notion, HubSpot, Google Drive, GitHub and more; personal or shared with the workspace | Docs say 1,000+; pricing page says 1,500+ |
| MCP servers | Point Lindy at a hosted server URL and it builds actions from the published tools | Public HTTPS streamable-HTTP only; per-connection availability and approval settings |
| Credentials | Save an API key once, share with the workspace, Lindy uses it without anyone pasting it into chat | Values cannot be read back; guardrails do not apply to credentials |
| Browser / computer use | Drives a browser when no integration or MCP server exists | Pro and Max per docs; see mismatch note |
| Pipedream-powered actions | 500+ extra actions across 200+ apps announced July 2025 | Changelog entry; current status not re-verified |
You are renting finished work by the credit, so you need a feel for what a "task" costs.
| What you are doing | What you pay / hit |
|---|---|
| A quick answer, summary or lookup | Lindy says 2 to 250 credits (its "Everyday asks" band) |
| A research report or a day of support triage | 250 to 1,000 credits ("Deep work") |
| A dashboard, a deck or an internal tool | 1,000 to 2,500 credits ("Big builds") |
| One Big build on the entry pool | A 2,500-credit job is 83% of one 3,000-credit seat for the month (JAVIS arithmetic), roughly $25 at the top-up rate of $10 per 1,000 |
| Running out mid-month | Lindy pauses credit-using actions. Admins can top up at $10 per 1,000, upgrade, or turn on Overages at 2x the regular credit cost. Paused tasks do not restart automatically |
| Letting one person drain the pool | Admins can set a per-teammate credit cap in Settings, Members (the Edit credit limit menu). Unset means unlimited draw |
| Unused monthly credits | Do not roll over. Purchased top-ups do not expire |
| An API for developers | I found no public Lindy API or SDK in the current docs index; developer extensibility is MCP, credentials and skills |
My practical read: the docs themselves say "most teams never come close to their limit", while the Trustpilot reviews describe people draining 3,000 credits in a day. Both can be true for different workloads. The way to find out is the free week: put Lindy on one realistic task, open the usage page, and extrapolate before you add seats. Remember a busy shared channel burns credits faster than any one person's DMs.
SOC 2 Type II and approvals on writes, with one gap you should know about: guardrails only cover shared Slack threads.
| Control | What Lindy says |
|---|---|
| SOC 2 | "SOC 2 Type II Certified", audited by Johanson Group (security page) |
| HIPAA | Security page: controls "mapped to HIPAA requirements". Pricing page: HIPAA with a signed BAA is Enterprise only. Homepage footer lists "HIPAA compliant" |
| GDPR and PIPEDA | Listed as compliant on the security and pricing pages |
| Training on your data | "Never sold, never shared, never used to train models" (vendor statement) |
| Encryption | In transit and at rest (vendor statement) |
| Approvals | Writes with outside impact wait for approval; reads from approved sources do not |
| Guardrails per integration | Always allow / Require approval / Don't offer, applying to writes only |
| Where guardrails do NOT apply | Web app chat, Slack DMs with Lindy, iMessage and SMS ("aren't guarded"), and saved credentials |
| Audit logs, SSO, SCIM | Enterprise. The security page says "Every action Lindy takes is logged" without naming the plan; the plan table puts audit logs on Enterprise |
| Meeting recording | Notetaker joins visibly; docs tell you to inform participants as your local laws require |
The practical read: the approval design is good where it applies, and the Slack-thread-only scope is a real nuance. If you let your personal Lindy loose on your own email by text, nothing asks you first.
Best for small Slack-centric teams that want an assistant, not an automation builder.
| Situation | Why Lindy fits | Likely plan |
|---|---|---|
| Small team in Slack that asks the same questions daily | A shared Lindy answers from approved sources and writes the report in the thread | Plus, pooled across seats |
| Founder or exec drowning in email and meetings | Daily brief, drafted replies in your voice, meeting recorder, follow-up bumps | Plus or Pro |
| Ops or growth lead who needs recurring reports | Routines on a schedule that post to a channel | Plus, watch routine frequency |
| Someone who needs computer use for odd web tools | Cloud browser for tools with no integration | Pro or Max |
| Regulated team needing a BAA, SSO or audit logs | Enterprise-only features | Enterprise |
| Team that wants deterministic, auditable step-by-step flows | Lindy is conversational and metered; classic builders fit better | Look at Zapier, Make or n8n first |
| Teams not on Slack | Teammate is built around Slack; iMessage, SMS, Gmail extension and web are secondary | Verify fit in the free week |
Six real gaps, not manufactured ones.
Lindy's own bands run from 2 to 2,500 credits per task and "deep research" is marked "Varies". Trustpilot reviewers describe a month of credits gone in a day.
Anyone who @mentions Lindy in Slack takes a seat, and removing one does not prorate. A busy channel can add cost without anyone deciding to buy.
The public changelog stops at Oct 21, 2025 and the pricing page and docs index do not explain the Agent Builder. A blog post dated Aug 31, 2026 still lists it and the docs Home screenshot has an Agents entry, so it seems to survive, but I could not confirm how agents are billed against the new credit pool.
Approval prompts apply in shared Slack threads, not in the web chat, DMs, iMessage or SMS.
Trustpilot shows 1.7 over 40 reviews with cancellation and support complaints. Small, self-selected sample, but consistent.
1,000+ versus 1,500+ integrations, Plus/Pro/Max versus Team naming, $30 versus $29.99 and $100 versus $99.99, and computer use on Team versus Pro and up. Small, but it means you should confirm at checkout.
Also: Skills is labelled early access, the Gmail extension is Gmail only, and the free week is not a standing free plan.
Lindy is best as a Slack assistant. Others win on determinism, price per run and developer control.
| If you mostly need... | Compare Lindy with... |
|---|---|
| Classic trigger-and-action automation across thousands of apps | Zapier review (live JAVIS review) |
| A visual scenario builder with cheaper entry pricing | Make review |
| Self-hostable, developer-friendly workflows with AI nodes | n8n review |
| Hosted integrations and developer plumbing for your own agent product | Pipedream review (note its Workflows shutdown, covered there) |
| Understanding MCP connectors before you wire any of these up | Claude MCP and Connectors guide |
| Building agents in code instead of renting a teammate | CrewAI review |
| Entry paid point (verified Oct 3, 2026) | Lindy | Zapier (per JAVIS review) | Make (per JAVIS review) | Pipedream (per JAVIS review) |
|---|---|---|---|---|
| Cheapest paid plan | $29.99/mo per user, 3,000 credits | Professional $19.99/mo billed annually | Core $9/mo | Startup $99/mo billed annually |
| What you pay for | Seats plus credits for finished AI work | Tasks | Credits | Connect plans (Workflows shuts down) |
| Free option | $50 credits for 7 days | Free plan | Free plan | Free plan |
No affiliate relationship shapes this review.
JAVIS has not established a publisher affiliate program with Lindy. Every CTA in this article points to the official product. If that changes, the review's facts, limitations and alternatives will not.
Go to the official page.
Open LindyRead the alternative that matches your real question.
Choose the next articleLindy sits between a chat assistant and an automation platform.
Every card below links to a published JAVIS article.
Questions people are actually searching right now
Is Lindy free?
Not as a standing plan. New accounts get $50 in credits that last 7 days, no card needed. After that, paid usage starts at $29.99 per user per month for 3,000 credits. Verified on lindy.ai/pricing, Oct 3, 2026.
How much does Lindy cost in October 2026?
$29.99 (3,000 credits), $99.99 (15,000) and $199.99 (35,000) per user per month on the self-serve tiers, with larger Team pools up to $9,999.50 for 1,750,000 credits and custom Enterprise pricing. The docs round these to $30, $100 and $200 and name them Plus, Pro and Max.
Did Lindy stop being an agent builder?
The front page, docs and pricing now describe Lindy Teammate, a Slack-native assistant. The agent builder does not appear to be gone: a Lindy blog post dated Aug 31, 2026 still lists "Agent Builder" as a feature, the app-builder and enterprise pages are live, and the docs Home screenshot has an Agents entry. But the public changelog's last entry is Oct 21, 2025, and the current pricing and docs index do not explain how agents are billed or documented, so ask support if you depend on them.
What happens when I run out of credits?
Lindy pauses credit-using actions until the next cycle and tells you. Admins can top up at $10 per 1,000 credits, upgrade, or enable Overages at 2x the regular credit cost. Monthly credits do not roll over; top-ups do.
Does Lindy do things without my approval?
Writes with outside impact (sending email, updating a ticket, posting, publishing) wait for approval in shared Slack threads. Reads do not. The docs say guardrails do not apply in the web chat, Slack DMs, iMessage or SMS.
Is my data used to train models?
Lindy says data is "never sold, never shared, never used to train models" and is encrypted in transit and at rest. That is a vendor statement; I did not audit it.
Does Lindy support MCP?
Yes, for hosted MCP servers with a public HTTPS streamable-HTTP endpoint. On connect it creates an action for every supported tool, and you can turn individual tools off.
What are the best Lindy alternatives?
Zapier, Make or n8n for deterministic automation, Pipedream for developer infrastructure, or a code-first framework such as CrewAI. See the alternatives table.
Where this came from, and when it was checked.
Pricing and plans: lindy.ai/pricing, checked Oct 3, 2026 including its FAQ; docs.lindy.ai pricing, teammate billing and Understanding Your Usage pages.
Product: lindy.ai homepage, docs.lindy.ai docs index and the full docs export (routines, skills, memory, meetings, iMessage and SMS, Chrome extension, integrations, MCP, credentials); lindy.ai sitemap (page dates); lindy.ai/changelog (entries from Mar 2025 to Oct 21, 2025); lindy.ai/blog/ai-teammate (dated Aug 31, 2026; calls the $29.99 tier the Plus plan and still lists Agent Builder); lindy.ai/enterprise and /app-builder (live).
Security: lindy.ai/security and docs.lindy.ai security page (vendor statements; no SOC 2 report was seen).
Market: Product Hunt page (4.6, 7 reviews, 1.4K followers); Hacker News items 35267091, 39385343 and comment search. G2 could not be read and Reddit was not available to us, so neither is cited as evidence; the G2 4.9 figure is attributed to Lindy's homepage only.
Real users: Trustpilot lindy.ai (1.7, 40 reviews, star split 7, 0, 5, 10 and 78 percent for 5 to 1 stars); the displayed score and the arithmetic average of the split (about 1.5) differ because TrustScore is Trustpilot's own weighted figure.
Alternatives and prices: published JAVIS reviews of Zapier, Make, n8n and Pipedream.
Official videos: youtube.com/watch?v=s3XEy4I35AA and m0bkNFIOLb4; author "Lindy" confirmed on YouTube.
Research method: every fact was read directly from official pages, docs and public APIs.
JAVIS Fit Score (66/100): editorial composite, not a benchmark: clear product direction and a coherent Slack-first experience (strong), approval and spend-cap controls (good, with the DM and iMessage gap), breadth of integrations and MCP (good), transparency of pricing (mixed: bands and credits are clear, outcomes are not), pricing and naming consistency across Lindy's own pages (weak), independent sentiment and billing trust (weak), low documentation of the earlier agent builder inside the new pricing (weak). Weighted by judgement; no automated test was run.
