Founder2 vs
the always-on agents
Always-on AI co-founders, checked on six things that matter. Every verdict is dated and sourced, ours included.
10
products compared
Always-on agents and AI co-founders
8
open to EEA individuals
Self-serve signup in the EEA
1
document EU residency
Processing or inference pinned to the EU, 1 more partly
3
gate actions on approval
A named human says yes first
Counts come from the checks below, as of November 2026.
Head to head
Pick any two. See the difference.
Price, models and the six capability checks side by side, plus where each product is genuinely stronger.
Pick ChatGPT dots for frontier-model general agency inside the ChatGPT ecosystem; pick Founder2 if you are an individual in the EEA, need a maintained plan document, or want metrics computed in code.
Founder2
from $29/mo
ChatGPT dots
from $100/mo (bundled into Pro tiers)
Where Founder2 is stronger
- Only product in this set with a living plan document the agent maintains and you approve.
- Only product in this set that computes unit economics (LTV, CAC, payback) in code from stored assumptions.
- Deterministic metrics from connected sources, not model-guessed numbers.
Checked 18 November 2026
Where ChatGPT dots is stronger
- Largest distribution in the category; bundled 'free' for existing Pro subscribers.
- Each dot gets its own cloud computer with a browser, plus 4,000+ app plugins.
- Reachable from Slack, Teams, iMessage and RCS.
Checked 18 November 2026
Every product, every check
Checked against each provider's own documentation and pricing pages. Founder2 gets the same marks as everyone else.
| Capability | Founder2from $29/mo | ChatGPT dotsfrom $100/mo (bundled into Pro tiers) | Grok Botfrom ~$30/mo (bundled into SuperGrok / Cursor plans) | Pokefrom $19/mo | Harkfree; pro2 from $20/mo | Instinctfree (invite-only beta) | Devinfree; Pro from $20/mo | Lindyfrom $29.99/mo | Manusfree; Pro from $20/mo | Martinfrom $21/mo (annual billing) |
|---|---|---|---|---|---|---|---|---|---|---|
| Available to an individual in the EEA | ||||||||||
| Documented EU data residency | ||||||||||
| A living business plan document | ||||||||||
| Dedicated persistent machine | ||||||||||
| Outward actions need named approval | ||||||||||
| Metrics computed by code |
- Yes, verified
- Partly, or with caveats
- No, or not offered
- Not verified yet
Founder2This is us
from $29/mo
- Available to an individual in the EEA
- Yes
- Documented EU data residency
- Yes
- A living business plan document
- Yes
- Dedicated persistent machine
- Yes
- Outward actions need named approval
- Yes
- Metrics computed by code
- Yes
ChatGPT dots
from $100/mo (bundled into Pro tiers)
- Available to an individual in the EEA
- No
- Documented EU data residency
- Not verified
- A living business plan document
- No
- Dedicated persistent machine
- Yes
- Outward actions need named approval
- Partly
- Metrics computed by code
- No
Grok Bot
from ~$30/mo (bundled into SuperGrok / Cursor plans)
- Available to an individual in the EEA
- Yes
- Documented EU data residency
- No
- A living business plan document
- No
- Dedicated persistent machine
- Yes
- Outward actions need named approval
- Yes
- Metrics computed by code
- No
Poke
from $19/mo
- Available to an individual in the EEA
- Yes
- Documented EU data residency
- No
- A living business plan document
- No
- Dedicated persistent machine
- No
- Outward actions need named approval
- No
- Metrics computed by code
- No
Hark
free; pro2 from $20/mo
- Available to an individual in the EEA
- Yes
- Documented EU data residency
- No
- A living business plan document
- No
- Dedicated persistent machine
- Yes
- Outward actions need named approval
- Partly
- Metrics computed by code
- No
Instinct
free (invite-only beta)
- Available to an individual in the EEA
- Not verified
- Documented EU data residency
- Not verified
- A living business plan document
- No
- Dedicated persistent machine
- Partly
- Outward actions need named approval
- No
- Metrics computed by code
- No
Devin
free; Pro from $20/mo
- Available to an individual in the EEA
- Yes
- Documented EU data residency
- Partly
- A living business plan document
- No
- Dedicated persistent machine
- Partly
- Outward actions need named approval
- Partly
- Metrics computed by code
- No
Lindy
from $29.99/mo
- Available to an individual in the EEA
- Yes
- Documented EU data residency
- No
- A living business plan document
- No
- Dedicated persistent machine
- No
- Outward actions need named approval
- Yes
- Metrics computed by code
- No
Manus
free; Pro from $20/mo
- Available to an individual in the EEA
- Yes
- Documented EU data residency
- No
- A living business plan document
- No
- Dedicated persistent machine
- Partly
- Outward actions need named approval
- Partly
- Metrics computed by code
- No
Martin
from $21/mo (annual billing)
- Available to an individual in the EEA
- Yes
- Documented EU data residency
- No
- A living business plan document
- No
- Dedicated persistent machine
- No
- Outward actions need named approval
- Partly
- Metrics computed by code
- No
Where each one wins, and where it does not
Strengths are what a competitor genuinely does better. The other column is where Founder2 is stronger or they fall short.
Founder2 vs ChatGPT dotsSuits Teams already on ChatGPT Business Premium who want always-on agents without leaving OpenAI's stack.
Pick ChatGPT dots for frontier-model general agency inside the ChatGPT ecosystem; pick Founder2 if you are an individual in the EEA, need a maintained plan document, or want metrics computed in code.
Where ChatGPT dots is stronger
- Largest distribution in the category; bundled 'free' for existing Pro subscribers.
- Each dot gets its own cloud computer with a browser, plus 4,000+ app plugins.
- Reachable from Slack, Teams, iMessage and RCS.
- Auto-review plus confirmation on sensitive actions (money, passwords) - an honest approval model.
- Specialist dots can hold credentials and narrow scopes - a genuinely good design.
Where Founder2 is stronger
- Individual Pro tiers with dots are not offered in the EEA, UK or Switzerland - only Business Premium is.
- OpenAI's EU inference residency for dots is reported but unverified; confirm before relying on it.
- No model picker - the model is OpenAI's choice.
- No living plan document and no deterministic business metrics.
- Bundled pricing makes the marginal cost of a dot opaque.
Founder2 vs Grok BotSuits US-based users who want powerful always-on agents with fine-grained approval controls and no EU residency requirement.
Pick Grok Bot for the strongest documented approval model and real computer use at a low price; pick Founder2 for EU residency, a plan document and deterministic metrics - Grok's computers and @bot surface are US-only.
Where Grok Bot is stronger
- Strongest documented approval model in the set: Auto Review classifier plus member-defined 'Ask first' rules, approval cards with Allow once / Always allow / Deny.
- Honest about its limits: docs state Auto Review 'does not review every side effect'.
- Persistent computer with durable disk, browser sessions and saved logins; acts on your local machine via the desktop app.
- Enterprise-legible: per-user isolation, SSO/SCIM, audit logs, action recording.
- Low entry price relative to frontier agent products.
Where Founder2 is stronger
- US-only data residency, stated plainly in docs - no EU option; @bot tagging is blocked across the EU, UK, Australia, Iceland, Liechtenstein and Norway.
- One computer per member shared by all their bots - bots see each other's sessions and files.
- Bots act as the signed-in member; blast radius equals the human's access.
- No model picker; the model mix can change with no guaranteed vendor set.
- No living plan document and no deterministic business metrics.
Founder2 vs PokeSuits Individuals who want an assistant reachable over iMessage, WhatsApp or Telegram without installing anything.
Pick Poke for the easiest possible start - a text thread that does tasks; pick Founder2 if you need a persistent workspace, a maintained plan, or documented approval gates before outbound actions.
Where Poke is stronger
- Best onboarding in the category: no app, no UI, just the messaging thread you already use.
- Available in all regions worldwide per vendor docs, with EEA/UK/CH privacy disclosures.
- Email, calendar, reminders, web search, integrations, recipes plus an API/MCP surface for builders.
- Ultra's human-in-the-loop task service is a differentiator nobody else offers.
Where Founder2 is stronger
- No documented persistent machine or durable desktop - anything needing a logged-in session must be rebuilt.
- Text-first is a ceiling: no workspace, no documents, no dashboard, no plan.
- No documented approval step before sending email or other outward actions.
- Models are opaque; no EU residency despite EEA availability (US-first processing).
- No living plan document and no deterministic business metrics.
Founder2 vs HarkSuits Consumer users who want one chat thread that acts on their behalf, not founders running a company workspace.
Pick Hark for a polished single-thread consumer assistant with a persistent computer; pick Founder2 if you need a business plan with metrics - and note Hark's Memory is disabled in the EEA, UK and Switzerland.
Where Hark is stronger
- Cleanest consumer experience in the set: one thread, no configuration, works without a manual.
- Unusually generous, legible pricing - no feature paywalls.
- Cloud computer per account; files and logins persist across sessions.
- Zero-retention third-party inference is a credible privacy posture.
- Asks before sending messages or making purchases unless that category is authorised.
Where Founder2 is stronger
- Memory - the feature that makes it 'always-on' - is disabled in the EEA, UK and Switzerland.
- No EU residency: transfers under standard contractual clauses; an observed EU region is config, not a commitment.
- Approvals are category-level, not action-level with payload review.
- No workspace, no multi-artifact plan, no economics - documents are produced on request.
- Privacy policy is JS-rendered and resisted automated retrieval, which makes diligence harder.
Founder2 vs InstinctSuits Early-adopter consumers comfortable with invite-only access and minimal published detail.
Pick Instinct if you can get an invite and want an assistant that operates real consumer apps end to end; pick Founder2 for a purchasable product with documented security, approvals and business metrics.
Where Instinct is stronger
- Most ambitious interaction model: uses consumer apps through their real interfaces, not APIs.
- Strong funding and talent signal - $2.5B Series B, reported ~$10B valuation.
- Free during beta; reached by text, call, iMessage and WhatsApp.
- Reportedly controls its own phone number and computer.
Where Founder2 is stronger
- Invite-only - not a purchasable alternative, and EEA availability is unresolved (reports conflict).
- No published security or approval model; reporting describes actions taken without explicit permission.
- Terms reportedly allow training on user data - disqualifying for most business use.
- Persistence across sessions is unverified.
- No documents, no plan, no metrics; opaque pricing roadmap.
Founder2 vs DevinSuits Engineering teams that want to delegate scoped coding tasks and review the resulting pull requests.
Pick Devin for autonomous software engineering against a repo; pick Founder2 if you need strategy, economics and a maintained plan rather than a coding agent.
Where Devin is stronger
- Genuinely productive on scoped engineering work; the pull request is an elegant approval gate.
- API exposes waiting_for_approval and bypass_approval states.
- Real EU residency - a Frankfurt cluster - but only on Enterprise plans.
- Transparent per-session ACU accounting.
Where Founder2 is stronger
- An engineer, not a co-founder: no strategy, economics or go-to-market.
- Session VMs mean nothing runs while you sleep unless you schedule it; state lives in the repo, not the machine.
- EU residency is paywalled behind Enterprise; self-serve residency is unverified.
- Pricing documentation contradicts itself on Teams billing.
- No living plan document and no deterministic business metrics.
Founder2 vs LindySuits Teams building multi-step AI workflows across many business tools.
Pick Lindy for team-scale workflow automation with the clearest approval gate and thousands of integrations; pick Founder2 if you are a solo founder who needs a plan document and deterministic unit economics.
Where Lindy is stronger
- Clearest named approval gate in the set: 'anything with outside impact waits for approval' - sending email, updating a ticket, posting - with a named approver.
- Model choice per task is real, not a marketing line.
- 3,000+ integrations plus MCP, Slack-native invocation, iMessage and phone.
- Unusually complete enterprise controls for a company this size.
Where Founder2 is stronger
- Credit pricing is impossible to forecast - you cannot tell a founder what a month costs until they have used one.
- Built for teams; a solo founder pays for a platform.
- No hosted desktop, so anything requiring a logged-in UI is out of reach.
- No EU residency option on self-serve or Team plans (US hosting).
- No living plan document and no deterministic business metrics.
Founder2 vs ManusSuits Users who want to hand off a defined task and get a finished artifact back.
Pick Manus for one-off deep tasks that produce deliverables - research, builds, slides; pick Founder2 if you need an ongoing plan and metrics rather than per-task sessions.
Where Manus is stronger
- Best 'give it a task, get a deliverable' experience: real browser, real files, real output.
- Breadth of model routing across frontier providers.
- Explicit per-task credit accounting - cost is visible.
- Optional always-on Cloud Computer is sold alongside the subscription.
- Projects and generated websites persist between sessions.
Where Founder2 is stronger
- Session-shaped, not founder-shaped: no plan, no P&L, no goals-and-metrics artifact.
- VM per task plus a paid add-on - not a persistent workspace machine by default.
- 'My Computer' asks before each local command, but cloud tasks run autonomously - wrong default for anything touching money or customers.
- No EU residency (processing in US and Singapore under SCCs).
- Credit math still surprises people.
Founder2 vs MartinSuits Individuals drowning in email and calendar who want a low-cost assistant that drafts and schedules.
Pick Martin for reliable inbox and calendar triage over text or call; pick Founder2 if you need more than personal-assistant scope - a plan, metrics and approval-gated business actions.
Where Martin is stronger
- Narrow and therefore reliable: inbox and calendar triage done well.
- Text, call, email and Slack access is low-friction.
- Drafts replies for approval before sending.
- Aggressive annual pricing ($21 vs $35 monthly).
- Available in the EEA on indirect evidence (DE/UK App Store listings).
Where Founder2 is stronger
- No computer, so no browser work; integration-based only (up to 5 Gmail + 5 Outlook accounts).
- Whether proactive sends are gated is undocumented - 'drafts for approval' is not 'cannot send without approval'.
- US-hosted with no EU option; no SOC 2 report and stale legal docs (2024).
- Model routing is opaque.
- No living plan document and no deterministic business metrics.
Pricing
What the entry price buys
The cheapest tier each product publishes, as of the check date. Annual billing and usage overages are not reflected.
Founder2
This is usLive data from Stripe, GitHub, Sentry, PostHog and Slack is included on paid tiers, and every plan gets a one-click catalog of about 90 tool servers over MCP. See the full pricing
Plus
$29/mo
30 VM hours · $15 coding pool · 1 seat · open-weight models
Pro
$99/mo
150 VM hours · $75 coding pool · 10 seats · closed frontier models
- Poke
Free
forever; no card required
- Hark
Free
every feature included; upgrades buy usage, not capability
- Instinct
Free
invite-only beta; no subscription 'for now'
- Devin
Free
light quota
- Manus
Free
300 credits/day capped at 1,500/month, 1 concurrent task
- Martin
$21/mo
billed yearly; $35/mo monthly
- Grok Bot
~$30/mo
bundles Grok Bot; higher tiers $100 and $200
- Lindy
from $29.99/mo
$50 of credits for 7 days
- ChatGPT dots
$100 / $200 / $500 per mo
dots bundled into top individual tiers; not on Plus or Go
"Custom" means there is no public price. Prices checked November 2026; confirm with each vendor before you buy.
How to read this
Every verdict was checked against the provider's own documentation and pricing pages, and each product carries a checked date. If a mark looks wrong, the sources are linked on each card. Send corrections to support@founderos.app.
We build Founder2, and it is one of the products on this page. We mark our own gaps with the same scale we use for everyone else. Not verified means we could not confirm either way, not that the feature is absent.
Longer reads
- Grok Bot, ChatGPT dots, and EU residencyThe dated, sourced deep-dive behind two rows of this matrix.
- AI agent approvals for foundersHow to review a proposed action before you let it through.
- What is an AI co-founder?The category defined plainly: assistant, agent or something else.
- What Founder2 can and cannot do todayThe capability note our own column is checked against.
- Welcome to Founder2The launch post: what it is, what it costs, where it stops.
See what $29/mo buys
A living plan, code-computed metrics and an approval-gated co-founder, plus an Agent VM for code.