The Rundown AI homepage

Independent tool overview

Lindy at a glance

Lindy is an AI teammate and agent platform that works through Slack, iMessage or SMS, Gmail, and its web app. It can research, draft files and messages, manage inbox and meeting workflows, run scheduled routines, use connected apps or MCP servers, and operate browser sessions. Current plans start at $29.99 per user per month for 3,000 credits, with team credits pooled by workspace. Lindy's ability to act across business systems is the appeal and the risk: teams need least-privilege connections, named approvers, prompt-injection defenses, meeting consent, and close review of every consequential action.

Visit the official Lindy site ↗
Lindy product preview
Product type
AI teammate, executive assistant, meeting assistant, and cross-app agent platform
Primary surfaces
Slack, web, Gmail, iMessage, SMS, and supported meeting platforms
Integrations
Lindy advertises 1,000+ integrations plus MCP connections
Built-in skills
More than 40, including research, data analysis, decks, dashboards, and team catch-ups
Meeting platforms
Google Meet, Zoom, and Microsoft Teams are named on the current site
Slack visibility
Public channels after workspace installation; private channels only when invited; personal DMs are excluded by the current FAQ
Approvals
Current teammate pages say outside-impact actions wait for a named approver
Computer use
Cloud browser sessions can retain logins and session data; documentation says sessions persist for 30 days after the last action
Model choice
The current pricing page says most major models are available
Training claim
Lindy says neither it nor its providers use customer data to train AI models
Compliance claims
SOC 2 Type II, GDPR, HIPAA, and PIPEDA; obtain evidence and a signed BAA where applicable
Starting price
$29.99 per user per month for 3,000 credits
Trial
A first-time Slack teammate gets seven days; current pricing says direct signups are billed immediately
Reviewed
August 31, 2026 from current product, pricing, documentation, security, privacy, and terms pages

Overview

What Lindy is

Lindy has evolved from a no-code custom-agent builder into a broader AI teammate positioned inside a company's daily work. An admin can add it to Slack, employees can ask for work in a thread or private assistant conversation, and the same service can operate through email, web, iMessage, SMS, meetings, schedules, and connected tools.

The current product combines conversational delegation with more than 40 built-in skills for tasks such as research, data analysis, decks, dashboards, meeting work, and reporting. Users can demonstrate a repeatable process and save it as a team skill, while scheduled routines handle recurring briefs, reports, follow-ups, and monitoring.

Lindy advertises more than 1,000 integrations plus MCP support. It can read approved company sources, update systems such as CRM or ticketing tools, create documents, send communications, and use a cloud computer when no direct integration exists. That breadth makes permission design, tool scoping, test data, and rollback more important than the prompt itself.

The current teammate interface says actions with outside impact—such as sending an email, changing a ticket, posting to another channel, or publishing a document—wait for a named approver, while read-only lookups from approved sources proceed without approval. Custom-agent documentation also exposes confirmation toggles, draft modes, and condition-based escalation. Verify the behavior of every action class rather than assuming an approval will catch an ambiguous side effect.

Lindy says data is encrypted in transit and at rest, is not sold, and is not used by Lindy or its providers for model training. The company advertises SOC 2 Type II, GDPR, HIPAA, and PIPEDA controls; Enterprise adds a signed BAA, SSO, SCIM, and audit logs. Buyers should review the actual report, DPA, current subprocessors, retention, model routing, data residency, and integration-specific scopes before connecting sensitive systems.

Use cases

Who Lindy is best for

The strongest fit depends on the job you need the product to complete, not the size of its feature list.

Slack-centered teams

Organizations that want employees to delegate work through familiar channels instead of learning a separate workflow builder.

Inbox-heavy professionals

Operators who need triage, draft replies, scheduling, reminders, and meeting follow-through across several accounts.

Meeting-intensive teams

Groups that need visible recording, transcripts, notes, searchable context, controlled sharing, and follow-up workflows.

Recurring operations

Teams producing scheduled briefs, reports, customer-success handoffs, CRM updates, and routine cross-app checks.

Cross-system research

Users who need answers synthesized from approved email, meeting, document, CRM, analytics, and public-web sources.

Governed agent pilots

Organizations prepared to use least privilege, test environments, named approvals, monitoring, and rollback before scaling autonomy.

Capabilities

Core Lindy features

1

Slack teammate

Responds in channel threads, supports private assistant conversations, and works from approved team sources.

2

Personal assistant

Delegates work through the web app, iMessage, or SMS using personal email, calendar, meeting, and preference context.

3

Inbox management

Triages and prioritizes email, drafts replies in the user's voice, and can schedule or send once approved.

4

Meeting lifecycle

Prepares briefings, joins named meetings, records and transcribes them, generates notes, tracks actions, and drafts follow-ups.

5

Shared meeting library

Organizes recordings and transcripts in folders with owner-controlled visibility for individuals or teams.

6

Scheduled routines

Runs recurring work such as daily briefs, weekly reports, reminders, catch-ups, and operational checks.

7

Built-in skills

Provides more than 40 skills spanning research, analysis, presentations, dashboards, communication, and routine work.

8

Custom team skills

Lets a demonstrated process be saved and reused as a skill across the workspace.

9

Persistent, editable memory

Stores workspace or personal context in plain files that users can inspect and change.

10

Connected actions

Reads and acts across more than 1,000 supported services, including productivity, CRM, support, data, and developer tools.

11

MCP support

Connects compatible Model Context Protocol servers to extend access beyond native integrations.

12

Computer use

Operates websites through a persistent or incognito cloud browser when an API integration is unavailable.

13

Model selection

Allows a user or workflow to select among supported major AI models according to task needs.

14

Human approval

Supports named approvers, confirmation steps, draft-only actions, and condition-based escalation for consequential work.

15

Credit controls

Pools plan credits across the workspace, pauses work when depleted, and lets admins set allocations or purchase top-ups.

16

Enterprise controls

Adds SSO, SCIM, audit logs, signed-BAA support, dedicated help, and onboarding according to current public pages.

Process

How the Lindy workflow works

  1. Step 1

    Pick one bounded job

    Start with a repetitive, measurable workflow whose inputs, outputs, owner, acceptable error rate, and rollback are already understood.

  2. Step 2

    Classify the data and impact

    Identify personal data, customer secrets, regulated records, credentials, financial exposure, external communications, and destructive actions.

  3. Step 3

    Choose the least-powerful route

    Prefer a read-only native integration over write access, MCP, or computer use; avoid giving one agent broad access across unrelated systems.

  4. Step 4

    Create a dedicated test boundary

    Use sandbox accounts, test channels, non-production records, dedicated browser sessions, and a small representative dataset.

  5. Step 5

    Connect minimum scopes

    Authorize only required inboxes, calendars, Slack channels, folders, records, tools, and actions; exclude private or regulated sources.

  6. Step 6

    Write explicit operating rules

    Define the goal, allowed sources, prohibited actions, uncertainty behavior, output format, escalation thresholds, and completion criteria.

  7. Step 7

    Require named approval

    Keep approval on for every external message, record change, file publication, financial step, deletion, scheduling commitment, and other consequential action.

  8. Step 8

    Test hostile and ambiguous inputs

    Include prompt injection, malicious attachments, misleading emails, duplicate events, missing fields, revoked permissions, timeouts, and conflicting instructions.

  9. Step 9

    Verify source and destination

    Before approval, check that Lindy used the right tenant, account, record, people, dates, files, factual sources, and current version.

  10. Step 10

    Watch the first live runs

    Review traces, approvals, token or credit consumption, integration errors, duplicate actions, and downstream system changes.

  11. Step 11

    Set spending and failure limits

    Cap credits by person, limit loops and batch size, add stop conditions, pause on anomalies, and define who responds to an incident.

  12. Step 12

    Expand one permission at a time

    Increase autonomy only after the bounded workflow meets accuracy, safety, latency, cost, and recovery targets over enough real runs.

  13. Step 13

    Audit continuously

    Review connected accounts, workspace membership, meeting sharing, browser sessions, memory files, action logs, spend, retention, and unused credentials.

Cost

Lindy pricing and free plan

Lindy's current public pricing is per user and credit based. Plus is $29.99 per user monthly for 3,000 credits, Pro is $99.99 for 15,000, and Max is $199.99 for 35,000. Every seat contributes to a shared workspace pool; unused credits do not roll over, and credit-using work pauses when the pool is empty. The current homepage says admins can top up at $10 per 1,000 credits. First-time users who join through Slack receive a one-week trial, while direct signups are billed immediately. Older documentation still shows a prior $49.99 Plus price and different feature gates, so use the live pricing page and checkout as authoritative.

Plus

$29.99/user/month

For everyday assistant usage.

  • 3,000 credits per user per month
  • Credits enter the workspace's shared pool
  • Current pricing lists Slack, routines, memory, skills, meetings, inbox work, computer use, integrations, MCP, model selection, and approvals
  • Credit usage can range from a few credits for simple asks to thousands for large builds
  • Unused credits do not roll over
  • Current public pages do not show an annual price

Pro

$99.99/user/month

For heavier usage and more connected inbox capacity.

  • 15,000 credits per user per month
  • Five times the Plus credit allocation
  • Credits are shared within the workspace pool
  • Current homepage says Pro and Max add more connected inboxes but does not publish the exact per-tier count
  • No permanent free plan is shown
  • Confirm current feature gates and billing terms at checkout

Max

$199.99/user/month

For the highest public self-serve usage tier.

  • 35,000 credits per user per month
  • Roughly twelve times the Plus credit allocation
  • Credits are shared within the workspace pool
  • Designed for heavy recurring workflows and large builds
  • Work pauses when the pool is exhausted unless more credits are added
  • Confirm workspace, cancellation, seat, and top-up rules before rollout

Enterprise

Custom quote

For organizations requiring stronger identity, audit, compliance, and enablement controls.

  • Everything in Max
  • SSO and SCIM listed on the current homepage
  • Audit logs
  • HIPAA support with a signed BAA
  • Shared usage and bonus credits
  • Dedicated support, onboarding, and enablement
  • Request SOC 2 report, DPA, subprocessor list, data-flow diagram, retention, residency, incident terms, and service levels

Pricing checked . Check current pricing at the source ↗

Assessment

Lindy strengths and limitations

Where it stands out

  • Meets teams inside Slack, email, meetings, text messaging, and the web instead of forcing every user into a new builder
  • Combines assistant-style delegation with reusable skills and scheduled automation
  • More than 1,000 integrations and MCP support cover a broad operational surface
  • Built-in source and meeting context can make research more relevant to the company
  • Editable memory files give users more visibility than an opaque personalization layer
  • Named approval for outside-impact actions is a strong default if consistently enforced
  • Shared meeting folders provide deliberate visibility controls
  • A shared credit pool lets heavy and light users draw from the same workspace allocation
  • Current pricing explains typical credit ranges and pauses work instead of silently generating an overage
  • SOC 2 Type II and Enterprise identity, audit, and signed-BAA options support formal review

What to consider

  • Broad access to email, calendars, Slack, meetings, CRM, files, and other tools creates a large blast radius if credentials, permissions, or instructions are compromised
  • Prompt injection can arrive through webpages, emails, documents, tickets, Slack messages, meeting transcripts, MCP tools, and integration outputs
  • Read-only lookups can still expose confidential data or combine information across contexts that should remain separated
  • An approval screen is only useful if the approver sees the exact final action, destination, data changes, and source evidence
  • The teammate page promises approval for outside-impact actions, while custom-agent documentation exposes configurable confirmation; test each route and action
  • AI-generated emails, reports, summaries, decks, and record updates can be incomplete, fabricated, misattributed, or inappropriate
  • Meeting recording and transcription require clear participant notice, lawful consent, retention, access, and deletion rules
  • Computer use retains website sessions and may store screenshots; persistent logins increase credential and cross-task contamination risk
  • Browser automation can violate a third-party site's terms, trigger bot defenses, duplicate actions, or lock an account
  • MCP extends Lindy's authority into separately operated servers whose security, schemas, logs, and prompt-injection handling must be reviewed
  • Model selection can route company data to different model providers and regions; confirm current processors and enterprise controls
  • The public privacy policy was last updated in February 2024 and may not fully describe the current teammate, meeting, MCP, model, and memory product surface
  • The generic privacy policy and public security claims are not substitutes for the current SOC 2 report, DPA, subprocessors, retention schedule, and architecture review
  • HIPAA suitability requires an eligible Enterprise arrangement, a signed BAA, correct configuration, minimum-necessary access, and the customer's own compliance program
  • Credits vary widely with task complexity, model, duration, and premium actions, so cost is hard to predict without a real workload trial
  • Credits expire each cycle, direct signups are billed immediately, and every active user takes a paid seat
  • Current pricing and older documentation conflict on the Plus price and some entitlements; confirm the live checkout and contract
  • Autonomous routines can repeat a bad assumption at scale unless limits, monitoring, idempotency, and rollback are designed first

Compare

Lindy alternatives

The right alternative depends on the specific output, workflow, controls and budget your project requires.

Agents

Zapier Agents

Consider Zapier Agents when app automation breadth and the existing Zapier ecosystem matter more than a Slack-native teammate experience.

Explore Zapier Agents

Business Operations

Gumloop

Consider Gumloop for visual, no-code AI workflows with explicit flow construction and data-processing steps.

Explore Gumloop

Business Operations

n8n

Consider n8n when technical teams want more workflow-level control, self-hosting options, and explicit orchestration.

Explore n8n

Questions

Lindy FAQs

What is Lindy?

Lindy is an AI teammate and agent platform that can research, create work, manage inboxes and meetings, run scheduled routines, and act across connected business tools from Slack, web, email, iMessage, or SMS.

How much does Lindy cost?

As of August 31, 2026, Plus is $29.99 per user per month with 3,000 credits, Pro is $99.99 with 15,000, Max is $199.99 with 35,000, and Enterprise is custom.

Does Lindy have a free plan?

No permanent free plan is listed. The current site says a new teammate who first joins through Slack gets seven days before billing, while a direct signup is billed immediately.

How do Lindy credits work?

Each seat contributes credits to a shared workspace pool. Work can cost from a few credits to thousands depending on the model and task. Credits reset without rollover, and credit-using actions pause when the pool is depleted.

Why do some Lindy pages show $49.99 for Plus?

Some documentation still reflects an older pricing structure. The live pricing page reviewed on August 31, 2026 lists Plus at $29.99 per user per month. Confirm the final price and entitlements at checkout.

What can Lindy access in Slack?

The current FAQ says Lindy can see public channels after installation and private channels only when invited by a channel manager. It says archived channels and individuals' personal DMs are excluded.

Will Lindy send or change things without approval?

The current teammate pages say outside-impact actions wait for a named approver. Because custom flows and integrations can behave differently, test every action and retain approval for messages, record changes, publications, money, deletions, and commitments.

Can Lindy record meetings?

Yes. Lindy can join Google Meet, Zoom, and Microsoft Teams as a visible participant, record and transcribe the meeting, create notes, and organize the result. The organization remains responsible for notice, consent, access, retention, and deletion.

Does Lindy support MCP?

Yes. Lindy advertises connections to any compatible MCP server. Treat each server as another privileged integration and review its operator, tools, permissions, data flows, authentication, logs, and prompt-injection defenses.

What is Lindy Computer Use?

Computer Use lets an agent operate websites through a cloud browser when no API integration is available. Persistent computers can retain logins and session data, while incognito computers do not retain the session; documentation says persistent sessions remain for 30 days after the last action.

Does Lindy train AI on customer data?

Lindy says neither it nor its providers use customer data to train models and that customer data is not sold. Buyers should confirm the current DPA, model providers, subprocessors, retention, human access, and zero-retention options.

Is Lindy SOC 2 compliant?

Lindy publicly claims SOC 2 Type II certification audited by Johanson Group. A serious buyer should request and review the report's entity, scope, period, exceptions, complementary controls, and bridge letter.

Can Lindy process protected health information?

Only consider PHI under an eligible Enterprise agreement with a signed BAA and verified configuration. A public HIPAA claim alone does not make every plan, model, integration, workflow, or customer use compliant.

What should I automate first in Lindy?

Start with a reversible, low-risk, read-heavy workflow such as a draft daily brief or internal report. Add narrow permissions, named approval, test data, stop conditions, monitoring, and rollback before external or destructive actions.

Is Lindy better than Zapier or n8n?

Lindy is strongest when teams want a conversational Slack-native teammate with meetings, email, memory, and delegated work. Zapier emphasizes a broad hosted automation ecosystem, while n8n offers more explicit technical orchestration and self-hosting options.

Bottom line

Our Lindy verdict

Lindy's current teammate direction is compelling for a Slack-centered organization: employees can delegate meaningful work in plain language, reuse shared skills, and combine email, meetings, files, research, schedules, integrations, MCP, and browser use without living in a workflow editor. The $29.99 entry plan is accessible, but useful deployment cost depends on variable credit consumption and paid seats. The platform should be adopted as a governed operator, not a trusted coworker by default. Begin with read-heavy work, keep consequential actions behind named approval, isolate credentials and browser sessions, test hostile inputs, and require evidence-backed security and privacy review before connecting regulated or high-impact systems.

Visit Lindy website ↗
The Rundown University

AI training for the future of work.

Get access to all our AI courses, hundreds of real-world AI use cases, live expert-led workshops, an exclusive network of AI early adopters, and more.

AI Courses

Get unlimited access to all of our current & upcoming industry-specific AI courses for the duration of your subscription.

Daily Guides

To keep up with the rapid pace of AI, our team publishes AI implementation guides daily. Our library contains 300+ practical use cases to automate real-world work.

Workshops

Join weekly, live, interactive sessions with industry leaders who are at the forefront of AI for hands-on implementation guidance and exclusive insights.

Community

Network with an exclusive community of AI-first professionals who are working smarter with AI. Learn how early adopters are using AI in their work and businesses.