The short version. Cursor at $20 is where the building happens. Supabase gives you a real backend free and $25 when it matters. PostHog has the most generous free tier in analytics by a distance, at a million events a month. Linear is free for one person and its AI features do not arrive until $16. Attio is the CRM to add the day you have more conversations than you can hold in your head.
Every price here came off the vendor’s own pricing page on 15 August 2026.
Running a product on your own? Our directory keeps shortlists for solo SaaS founders, covering the parts that are not the product.
What a company of one should actually pay for
The useful thing about this stack is how little of it costs anything until you have users. Four of the five here have free tiers that are not trials, and for a product without customers that is the correct amount to spend.
So we ranked on:
- How far the free tier goes. Not whether one exists, but whether it survives your first hundred users.
- Whether it works without a team. Most tooling assumes somebody else does the setup and somebody else reads the output.
- What it costs when it works. Usage-based pricing is kind at zero and occasionally alarming at scale, and it is worth knowing the shape before you depend on it.
- Whether it does a job you are avoiding. Founders buy tools for the work they enjoy. The value is in the work they do not.
1. Cursor: the fastest way to be two developers
Hobby tier free with limited agent requests. Individual $20/month, with Pro, Pro+ and Ultra tiers above it on agent limits. Teams $40/user/month. Verified 15 August 2026.
For a solo founder the argument is simpler than it is for a team. You have no colleague to ask, no reviewer, and nobody who knows the part of the codebase you wrote nine months ago and have forgotten. Something that understands the whole project and can make a change across several files is the closest thing to a second person you can buy for $20.
Where it earns most is unfamiliar territory. The Stripe webhook you write once, the migration you have not done before, the CSS that should be simple. Those are the tasks that eat a solo founder’s week and they are exactly what this is good at.
The Hobby tier is enough to find out whether you like it. If you already pay for Claude, Claude Code comes with a Pro subscription and covers similar ground from the terminal, so try both before adding a second bill.
Who it is wrong for: founders who are not writing the code. If you are non-technical and shipping through contractors, this is not your purchase.
2. Supabase: a real backend for nothing, then $25
Free plan with 500 MB database and shared compute. Pro $25/month with 8 GB database included and $10 a month in compute credits, which covers one Micro instance. Team $599/month. Enterprise custom. Verified 15 August 2026.
Postgres, auth, storage, realtime and edge functions, with a free tier you can genuinely launch on. For a solo founder that removes a week of setup and a category of decisions you are not qualified to make on your own, which is most of the value.
The pricing detail worth understanding is compute. Pro at $25 includes $10 of compute credits, which covers one Micro instance. Additional projects carry their own compute cost, so a founder running staging alongside production is not paying $25, and the surprise usually arrives in month two.
The jump from Pro to Team at $599 is enormous, which is fine because almost nobody reading this needs Team. Just know the ladder has a large gap in it.
Who it is wrong for: anyone whose product is not database-shaped, and anyone already comfortable running their own Postgres, who is paying for convenience they do not need.
3. PostHog: a million events a month, free
Usage-based with a monthly free tier that resets: 1 million analytics events, 5,000 session replays, 1 million feature flag requests, 1,500 survey responses, 1 million data warehouse rows, 100,000 error tracking exceptions and 500 PostHog AI credits. Beyond that, analytics is $0.00005 per event falling to $0.000009 at high volume, session replay $0.005 per recording, feature flags $0.0001 per request, surveys $0.10 per response, error tracking $0.00037 per exception. Verified 15 August 2026.
A million events a month free is a lot. Most pre-revenue products will not touch it, and the free allowance covers analytics, replays, flags, surveys and error tracking rather than one of them.
Session replay is the feature that matters most to a founder with few users. When ten people sign up and two get stuck, watching those two is more informative than any funnel chart, because the reason they left is usually something you would never have thought to measure.
The thing to watch is that usage pricing compounds across products. Each one is cheap per unit and there are seven of them, so the bill arrives as a sum you have not modelled. Turn on what you will look at and leave the rest off.
Who it is wrong for: founders who will not watch replays or read the numbers. This is a tool for people who will act on it, and installing analytics is not the same as using them.
4. Linear: free for one, and the AI is at $16
Free plan, which includes the agent platform and Linear Agent. Basic $10 per user per month billed annually. Business $16 per user annually, which is where Triage Intelligence, Code Intelligence in beta, and Linear Asks sit. Enterprise custom. Coding sessions and Loops require AI credits separate from the plan. Verified 15 August 2026.
Linear is the one people are surprised to see recommended for a company of one, since issue tracking sounds like a team problem. The reason it is here is that a solo founder’s real failure mode is holding everything in their head and then losing a week to whichever thing shouted loudest.
The free plan is genuinely enough for one person and it now includes the agent platform, which is unusual generosity at zero.
Note where the AI actually sits. Triage Intelligence and Code Intelligence are on Business at $16, not on Basic at $10. And coding sessions and Loops consume AI credits that are separate from your plan, so budget for that rather than assuming the seat price covers it.
Who it is wrong for: founders happy with a text file and a list. That genuinely works at this size, and adopting a process tool to feel organised is a way of avoiding the work.
5. Attio: when you stop remembering the conversations
Free tier with 100 seat credits and 250 workspace credits a month. Plus $44 per user per month, or $35 annually, with 500 seat credits. Pro $99 per user, or $79 annually. Enterprise custom. AI features including Ask Attio, agents and call intelligence are on every paid plan. Verified 15 August 2026.
Early on you do not need a CRM, and anyone who tells you otherwise is selling one. You need it at a specific moment, which is when you cannot remember whether you followed up with the person who asked about the enterprise plan three weeks ago.
Attio suits a founder because the data model bends to how you actually sell rather than assuming a standard pipeline, and because the AI is on every paid tier rather than reserved for the expensive one. The free tier is enough to find out whether you have reached the moment.
If your product is self-serve with no sales conversations at all, skip this and put the money into beehiiv or Kit instead, because your equivalent problem is an audience rather than a pipeline.
Who it is wrong for: pure self-serve products. There is nothing to track, and $35 a seat is a lot for a contact list.
Founders this list will not help
Funded startups with a team. Your constraints are coordination and hiring, and half of this is priced for one person on purpose.
Non-technical founders. Three of these five assume you write the code. Your stack looks different and starts with who is building.
Agencies and consultancies. You are selling time rather than a product, and the shortlists for that are genuinely different.
Anyone hoping tooling substitutes for talking to users. Every tool here helps you build and measure. None of them tells you whether anyone wants it, and the number of solo founders with an immaculate stack and no customers is the whole cautionary tale of this category.
Considered, and left off
Vercel and the hosting platforms. Genuinely part of most of these stacks. Left out because the choice is usually made by your framework rather than by comparison, and the free tiers are all adequate at this size.
Stripe. Not a decision. If you are charging, you are probably on it.
Notion. Excellent for the documents and the thinking, and covered in our lists for consultants and operations managers. For a solo founder it is optional in a way the five above are not.
beehiiv and Kit. The right fifth entry instead of Attio if your growth is audience-led rather than sales-led. Covered in our list for content publishers.
Questions founders ask us
What should I pay for before I have revenue?
Cursor at $20, and nothing else. Supabase, PostHog and Linear all have free tiers that will carry you past your first hundred users, and Attio is a purchase for a problem you do not have yet. Twenty dollars a month is a reasonable pre-revenue tooling budget.
Is usage-based pricing risky for a small product?
It is kind at zero and it compounds quietly. PostHog charges per event, per recording, per flag request and per exception separately, each cheaply, and the bill is the sum of seven small numbers. Enable what you will actually look at and leave the rest switched off.
Does Linear’s free plan include the AI?
It includes the agent platform and Linear Agent. Triage Intelligence and Code Intelligence are on Business at $16 a user. Separately, coding sessions and Loops consume AI credits that sit outside the plan price, so the seat cost is not the whole cost.
Do I need a CRM as a solo founder?
Not until you notice you have forgotten a conversation that mattered. That moment is a real signal and it arrives at different times for different products. Before it, a CRM is a tidy-looking way to avoid selling.
Cursor or Claude Code?
Try both. Cursor is the editor and suits people who want to see every change. Claude Code runs in the terminal, takes a task rather than a prompt, and comes with a Claude Pro subscription you may already pay for, which changes the comparison. They cost about the same and almost nobody needs both.
Will AI let one person run a real software business?
It has genuinely moved what one person can build and maintain, and that is not marketing. What it has not changed is distribution. The constraint on solo software businesses was rarely engineering capacity, and every tool here makes the buildable part faster without touching the part that decides whether it works.
See our maintained directory shortlists for solo SaaS founders, including the marketing site and the cash view.
Last reviewed: 15 August 2026. Every price was read off the vendor’s own pricing page on that date. We review this page quarterly.
Leave a Reply