The $0 AI stack isn’t a gimmick — it’s a real strategy. The constraint forces focus, and the focus is what gets the first paying customer. I ran 500k.io’s first 9 months on essentially this stack. Closed my first paid customer at month 4. Hit $1K MRR by month 5. The total tool spend across that whole period was $20 (Claude Pro) plus a $12 domain. Everything else: free.
I’m at $9,500 MRR / $114K ARR / 22.8% to my $500K target on 500k.io. Separately, I co-founded The Kreators AI with Jack — the agency manages about $45M of Meta Ads ($10M on my side, $35M on Jack’s). The agency runs on a much fatter stack. The solo project ran on this. Both work. Different stages.
The 5-tool $0 stack
| Tool | Function | Free-tier limit | What breaks first |
|---|---|---|---|
| Claude Free | AI assistant | ~30-50 messages/day | Daily message cap |
| Beehiiv Free | Newsletter | 2,500 subs | Subscriber cap |
| Cloudflare Pages | Hosting | Unlimited bandwidth | Build minutes (rare) |
| Resend Free | Transactional email | 3,000/mo, 100/day | Volume cap |
| Notion Free | Wiki + content org | 5MB upload | Upload size |
Total: $0/mo. Plus a domain ($12-20/year, the only non-negotiable spend).
That’s it. Five tools. Less than what most founders subscribe to in their first weekend.
Why this 5 and not the trendy 15
Every founder I know tried the 15-tool stack first. It failed not because the tools didn’t work — they all worked. It failed because the founder spent the first 3 weeks subscribing, integrating, and configuring instead of shipping.
The $0 stack works because:
- Onboarding is fast. All 5 tools in 90 minutes total.
- No mental overhead. Five tools means you remember what each does.
- Focus is forced. You can’t lean on a paid feature you don’t have.
The hidden math: subscribing to 15 tools costs ~12 hours of setup time across the first month. That’s 12 hours not spent finding customers.
Tool 1 — Claude Free (the brain)
The free tier of Claude.ai gives you Claude 4.6 Sonnet (current default, May 2026) with a daily message cap. You can’t run Claude Code on free, but you can:
- Write articles
- Research topics
- Edit drafts
- Plan strategy
- Audit landing pages
Daily message cap: somewhere between 30-50 (Anthropic doesn’t publish exact limits and they vary). For a founder shipping 1-2 articles per week, it’s enough.
Why Claude over ChatGPT free: better long-form writing, better instruction-following, fewer mid-day model swaps to weaker variants on the free tier.
Upgrade trigger: the day you start hitting the message cap before your day’s done. For most founders, this happens around month 3-4 if they’re actually shipping daily content.
Tool 2 — Beehiiv Free (the asset)
Free tier: 2,500 subscribers, unlimited sends, custom domain, basic analytics. No automations.
The 2,500-subscriber cap is the single most important threshold in the whole $0 stack. Everything else has roomy ceilings. Beehiiv Free will run out before anything else.
Why Beehiiv over alternatives:
| Platform | Free-tier subs | Custom domain | Verdict |
|---|---|---|---|
| Beehiiv | 2,500 | ✓ | Best free tier |
| Substack | Unlimited | Limited | Takes 10% revenue, brand on their domain |
| ConvertKit (Kit) | 1,000 | ✓ | Tighter free tier |
| Mailchimp | 500 | No | Worst free tier in 2026 |
Beehiiv is the only platform where the free tier scales to where solo founders actually need to go.
Upgrade trigger: 2,000 subscribers (so you have buffer for the migration / Scale tier setup).
Tool 3 — Cloudflare Pages (the host)
Free tier: 500 builds/month, unlimited bandwidth, unlimited requests, custom domain, automatic HTTPS, edge caching globally.
There is genuinely no reason to pay for hosting until you’re running an enterprise-scale app. Cloudflare Pages is best-in-class for static sites, Astro sites, Next.js sites, and most SSR setups.
The build cap (500/month) sounds restrictive but is generous. You’d need to push 16+ builds per day to exceed it. Most founders push 3-8 per day.
Upgrade trigger: literally never for content sites. Workers (the serverless function layer) bills past 100K requests/day, but that’s a separate product.
Tool 4 — Resend Free (the email infrastructure)
Free tier: 3,000 emails/month, 100/day, 1 custom domain.
Resend handles welcome emails, password resets, magic links, receipts, lead-magnet delivery. Things that aren’t your newsletter but require sending email reliably.
Why Resend over alternatives:
| Platform | Free-tier volume | API quality | Verdict |
|---|---|---|---|
| Resend | 3K/mo | Best | Top pick |
| Postmark | 100 trial | Excellent | $50/mo too soon |
| SendGrid | 100/day | Mediocre | Outdated |
| Mailgun | 100/day | Old-school | Avoid |
| AWS SES | Unlimited (low cost) | Brutal setup | For developers only |
Resend’s free tier covers most pre-PMF apps. 3,000 emails/month is enough for ~500-800 active users with light email needs.
Upgrade trigger: when you’re consistently hitting the 100/day or 3,000/month ceiling. Usually around 1,000-2,000 active users.
Tool 5 — Notion Free (the wiki)
Free tier: unlimited pages, 5MB file uploads (annoying but workable), integrations, sync across devices.
Notion is where everything that isn’t on the website lives: client folders, content drafts, ideas, the editorial calendar, customer notes, contracts, the brief queue.
Free tier limitation that matters: 5MB upload cap kills PDFs over a certain size. Workaround: use Cloudflare R2 ($0 for 10GB free) for files, link from Notion.
Upgrade trigger: when you start needing teamspaces or Notion AI for note cleanup. $10/mo. Most solos can stay free for a year.
The sixth optional tool: a domain
$12-20/year for a .com or .io. This is the only thing on the list that costs real money. Pay it.
Buy from Cloudflare Registrar — they sell at-cost (no markup), and DNS is built-in. Cheaper and cleaner than Namecheap or GoDaddy.
What the $0 stack actually runs
I tested this exact stack on 500k.io for 9 months. Output during that period:
- 28 articles published
- 47 newsletter subscribers (kept under the cap intentionally for the test)
- 1 paying customer at $1,500/mo
- Live dashboard with Stripe data
- 4 lead magnets distributed via Resend
- Shipped weekly without missing
Total tool spend: $20 (Claude Pro). Plus the $12 domain. Total monthly OPEX: $20.
Output quality vs the $565/mo stack: roughly 80%. The 20% gap was speed (manual workflows took longer) and reach (no Perplexity meant slower research). But the substance was the same.
“Anyone telling you that you need a $500/mo stack to start a solo business in 2026 is wrong. They probably also haven’t started one. Start at $0. Add tools when free actually constrains output. Don’t subscribe to your future self.”
When the $0 stack breaks
Three signals tell you it’s time to upgrade.
Signal 1 — Beehiiv approaching 2,000 subs
Migrate to Scale ($49/mo) at 2,000-2,200 subs to give yourself buffer. Don’t wait for the cliff. The signup form will fail at 2,500 and you’ll lose ~3 days of new subs while you upgrade.
Signal 2 — Daily Claude Free message cap blocking shipping
If you’re hitting “rate limit” messages 3+ days/week and it’s blocking your work, upgrade to Claude Pro ($20/mo). The math: $20 / 30 days = $0.67/day. One blocked work session per month justifies it.
Signal 3 — First paying customer
The moment you cross $0 to $1 in revenue, your stack should match the seriousness of the business, not the seriousness of the experiment.
Add: Plausible $9 (real analytics), Stripe (free per-transaction), Calendly Free (booking).
You’re now at ~$30-50/mo of OPEX. Still cheap. Real business.
What the $0 stack is missing
Three categories of tool that aren’t here, and why:
1. Analytics beyond basic. Cloudflare Web Analytics is free and decent. Plausible at $9 is better. Skip GA4 unless you need it. The $0 stack uses Cloudflare’s free tier.
2. Scheduling / calendars. Calendly Free covers 1 event type, no payments, no team. For a founder taking 5-10 sales calls/month, that’s enough.
3. Project management. Notion replaces this. You don’t need Linear, Asana, or Trello at $0 MRR.
What it costs to ramp from $0 to $100/mo
| Stage | Tool changes | New cost |
|---|---|---|
| Start | $0 stack | $0 |
| Month 2 | Add Claude Pro $20 | $20/mo |
| Month 3 | Add Plausible $9 | $29/mo |
| Month 4 | Add Perplexity Pro $20 | $49/mo |
| Month 5 | Beehiiv Scale (when crossing 2K subs) $49 | $98/mo |
| Month 6+ | Optional Notion Pro $10 | $108/mo |
That’s the natural ramp. From $0 to $108/mo over 6 months, each upgrade triggered by an actual constraint. By month 6 you’re roughly at the $100/mo entry stack.
Internal links
- The $0 stack: 12 free tools for solopreneurs — the wider $0 list with non-AI tools.
- The $100/mo AI stack: entry tier for solopreneurs — the next tier.
- The full 500k.io stack — the today version at $565/mo.
- First $1K MRR with AI: the founder playbook — what to ship on the $0 stack.
- The honest math: $500K solo SaaS in 18 months — long-term trajectory.
- Beehiiv vs ConvertKit 2026: operator deep-dive — the platform that gates your $0 stack.
External sources
- Beehiiv pricing — the 2,500 free-tier cap.
- Cloudflare Pages limits — confirm the $0 ceiling.
- Resend pricing — the 3K/mo free tier.
- Anthropic — Claude pricing — current free vs Pro tier.
What to do this weekend
If you have nothing: buy a domain. Set up Cloudflare Pages. Deploy a one-page landing in 60 minutes. Put a Beehiiv signup form on it. Sign up for Claude Free.
You’re now running. Total cost: $12. Total time: 90 minutes.
Subscribe to a paid tool only when free actually breaks. Not before.
FAQ
Can a solopreneur really run a business at $0 in tools?
Yes — for 6-9 months. The free-tier ceiling is real: ~5,000 newsletter subscribers, ~1,000 web visitors per day, ~5 deep research queries per day. Past those thresholds, the $0 stack breaks. Plan for it.
What's the most-used $0 tool you'd recommend?
Claude Free. The free tier still runs Claude 4.6 Sonnet for limited daily messages. For founders writing 1-2 articles per week, it's enough. For founders doing daily content + research, upgrade to Pro at $20.
Why isn't ChatGPT Free on the list?
Because it's worse than Claude Free for the same workload. ChatGPT's free tier uses an older model on weekends, the message cap is tighter, and the interface lacks the project organization Claude has.
What breaks first when you grow past $0?
The Beehiiv 2,500-subscriber cap. That's the single hardest stop. Either upgrade to Beehiiv Scale ($49/mo) or migrate. Most solos hit this around month 4-7.
Is the $0 stack a real strategy or just a content gimmick?
Real strategy. I ran 500k.io's first 9 months on a $0 stack (with Claude Pro $20 added in month 3). Hit my first $1K MRR before crossing the $0 threshold. The constraint forced focus.
What if I need a domain?
Buy one. $12-20/year for a .com or .io. The $0 stack is about monthly recurring cost, not one-time spend. A domain is non-negotiable; everything else can wait.