Lindy

What it does

Writes code

YesLindy Build generates whole applications from a prompt; the Teammate product also writes its own integrations against a tool's API when none exists.Build apps as powerful as they are beautiful, with frontend, backend, databases, and integrations fully built in.lindy.ai/app-builder ↗

Debugs and fixes

YesLindy Build runs QA agents that exercise the app and repair what they break; no equivalent claim covers code you wrote yourself.Skip tedious debugging with QA agents that click through your app and fix broken code.lindy.ai/app-builder ↗

Writes tests

PartlyQA agents exercise the app in a browser; the vendor never claims it produces test code you keepSkip tedious debugging with QA agents that click through your app and fix broken code.lindy.ai/app-builder ↗

Ships it

PartlyBuilt apps are deployed to public lindy.site URLs, but anything with outside impact waits for an approverThe simplest way to build production-ready apps.lindy.ai/app-builder ↗

Done-for-you service

NoYou run the software: Lindy installs into your own Slack and works out of the box, and the only human service named is Enterprise onboarding and dedicated support.No workflows to build, no API keys, no training sessions. Add Lindy to Slack and it works out of the box. One person adds it, and the whole team gets…lindy.ai/enterprise ↗

Not established: Reviews code, Whole-codebase work

How much it does unattended

Runs autonomously

PartlyRead-only work and scheduled routines run unattended, but anything with outside impact waits for a named approver by default; an owner or admin can flip an integration to Always allow to remove the gate.Anything with outside impact waits for approval such as sending an email, updating a ticket, posting to another channel, publishing a doc.lindy.ai/pricing ↗

Multi-agent

PartlyLindy Build pairs the builder with separate QA agents working the same app; the Teammate product gives each person a private Lindy plus one shared team Lindy, and the vendor describes no orchestration between them.Skip tedious debugging with QA agents that click through your app and fix broken code.lindy.ai/app-builder ↗

Agent permissions

YesPer-integration guardrails offer Always allow, Require approval, or Don't offer, and they cover writes only; approval prompts land in the Slack thread with Approve and Deny buttons.Reads are never guarded. Checking your calendar or searching your email always just works. Guardrails only cover writes.docs.lindy.ai/integrations/overview ↗

Which models it runs on

Claude

PartlyVendor claims most major models are selectable but names none, so Claude support is not establishedPick your model of choice for each task. Most major models are available.lindy.ai/pricing ↗

GPT

PartlySame unnamed 'most major models' claim; GPT is not listed anywhere on the sitePick your model of choice for each task. Most major models are available.lindy.ai/pricing ↗

Model choice

YesYou pick the model per task from a dropdown and can set a default; Lindy otherwise chooses for you, and model choice is ungated on every plan including trial seats.Pick your model of choice for each task. Most major models are available.lindy.ai/pricing ↗

Not established: Open models

Where you use it

In your editor

NoThe surfaces the vendor enumerates are Slack, iMessage and SMS, a Chrome extension that runs inside Gmail, and the chat.lindy.ai web app; no IDE or editor plugin appears on the site or in the docs.Lindy runs inside Gmail. Triaged inbox, a morning brief, drafted replies, and a sidebar chat, all without leaving your tab.docs.lindy.ai/features/chrome-extension ↗

On the command line

NoLindy runs shell commands and code internally as part of a task, but no terminal surface is offered to the user; the only chat surfaces documented are Slack, iMessage/SMS, Gmail and the web app.Credentials are what Lindy reaches for when it runs shell commands and code.docs.lindy.ai/integrations/credentials ↗

In your pipeline

NoGitHub is one of the 1,000+ integrations, but no pull-request or CI surface is claimed in the enumerated surface listWorks wherever you workdocs.lindy.ai/index ↗

In a browser

Yeschat.lindy.ai is a hosted web app with its own chat, Files browser, Routines and Meetings pages; guardrails deliberately do not apply there, only in shared Slack threads.Guardrails apply in shared Slack threads only. Your web app chat, Slack DMs with Lindy, and iMessage and SMS aren't guarded.docs.lindy.ai/integrations/overview ↗

Whose machine it runs on

Self-hosted

NoLindy is delivered only as a hosted service on the vendor's own cloud; no self-managed, VPC or on-premise option is offered anywhere on the site.Lindy runs on enterprise-grade cloud infrastructure with hardened access controls. Your data doesn't mingle with anyone else's.lindy.ai/security ↗

Open source

NoNo repository or source availability anywhere on the site; the product is described only as vendor-run cloud infrastructureLindy runs on enterprise-grade cloud infrastructure with hardened access controls. Your data doesn't mingle with anyone else's.lindy.ai/security ↗

What it costs to run

How it meters

YesPer seat and per credit at once: seats run $29.99 to $199.99 per user per month, each adding 3,000 to 35,000 credits (about one cent each) to one pool, with top-ups at $10 per 1,000 and overages at 2x.You pay per user, and everyone on the workspace shares one pool of credits.docs.lindy.ai/pricing ↗

Free tier

PartlyThe free week exists only on the Slack path: someone new who @mentions Lindy in a connected workspace gets a week before their seat bills, once per person. Anyone signing up directly pays from day one.New teammates who join through Slack get their first week free, so your team can try giving Lindy real work before any seat is billed. Direct signups…lindy.ai/pricing ↗

MCP server

PartlyMCP is inbound only: point Lindy at any hosted MCP server over public HTTPS and it builds an action per supported tool, with per-person guardrails and workspace sharing. No Lindy-hosted MCP endpoint another agent could query is documented.Point Lindy at a hosted MCP server and it builds actions from the tools that server publishes.docs.lindy.ai/integrations/mcp ↗

Bring your own key

PartlyYou can store your own tool API keys as workspace credentials; no bring-your-own inference key or cloud is offeredSave an API key once, keep it for yourself or share it with the workspace, and let Lindy use it without anyone pasting it into a chat.docs.lindy.ai/integrations/credentials ↗

Not established: API access

Buying it for a team

A company can buy it

YesOne workspace subscription on a single bill, with roles of Member, Admin and Owner, admin-managed seats and an Enterprise tier sold through sales.Teammate is one workspace subscription: every seat is on a single bill and all credits pool together.lindy.ai/pricing ↗

Seat model

YesPer seat with no published floor or ceiling. Anyone who uses Lindy consumes one, including people who add themselves by @mentioning it in Slack; a removed seat runs to the end of the paid cycle with no proration.Anyone who uses Lindy takes a seat, whether they sign up directly, join your workspace, or @mention Lindy in Slack.lindy.ai/pricing ↗

Pooled budget

YesOne shared workspace balance, not a per-seat wallet: each seat contributes its tier's allocation and admins can cap an individual's draw from the Members table.Credits are pooled at the workspace level. Each seat adds its plan's allocation to one shared balance the whole team draws from, and admins can cap…docs.lindy.ai/pricing ↗

Admin controls

YesAdmins set which channels and folders Lindy may read, per-integration guardrails, per-person credit caps, who may approve external actions, and workspace toggles for meeting recording, email drafting and triage.Admins can control which features are available to users across the workspace. Toggle each feature on or off to manage access.docs.lindy.ai/start-here/team-setup ↗

Audit log

PartlyAudit logs are an Enterprise line item; the security and enterprise pages separately claim every action Lindy takes is logged and visible, which the plan table does not bear out for Plus, Pro or Max.Lindy is SOC 2 and GDPR compliant, and Enterprise adds SSO, audit logs, and HIPAA with a signed BAA.lindy.ai/pricing ↗

Single sign-on

PartlySSO and SCIM user provisioning are Enterprise-only, and the plan table marks both absent on Plus, Pro and Max. The vendor says 'access through your identity provider' but never names SAML or OIDC.SSO & SCIM Access through your identity provider, with user provisioning from a single source of truth.lindy.ai/enterprise ↗

What happens to your code

Opt out of training

YesNot an opt-out but a blanket commitment on every tier: the vendor states your data is never sold and never used to train models, its own or its providers'.Lindy does not sell your data. Lindy and our providers do not use your data to train AI models. Clear policies govern how information is collected,…lindy.ai/security ↗

Data residency

PartlyThe vendor says residency requirements are supported but names no region, jurisdiction or hosting provider anywhere on the site or in the docs, and offers no self-hosted alternative.You control how your data is processed and stored. Data deletion requests, access rights, and residency requirements are all supported.lindy.ai/security ↗

Getting out

PartlyPlan credits expire at each cycle and never roll over, while purchased top-ups carry over indefinitely. On data, deletion requests are supported and Files is an editable versioned store you can open, but no bulk export is documented.Your plan's credits refresh at the start of each billing cycle and don't roll over, so every month starts fresh. Purchased top-up credits are the…docs.lindy.ai/pricing ↗

Certifications

YesSOC 2 Type II audited by Johanson Group, plus GDPR, HIPAA and PIPEDA, with reports available through a Drata trust centre. A signed HIPAA BAA is Enterprise-only.SOC 2 Type II, GDPR, HIPAA, and PIPEDA, encrypted in transit and at rest. Your data is never sold and never used to train a model, ours or anyone…lindy.ai ↗