Calculators and money-honest breakdowns for founders who run the numbers.
Featured interactive tool
Model your own monthly bill across OpenAI, Anthropic, and Google — or price a custom model. No signup.
Featured interactive tool
See your blended processing rate across a real transaction mix — not just 2.9% + 30¢. Average ticket, international cards, ACH, and more. No signup.
48 posts
Supabase Pro is $25 in 2026, but a growing SaaS paid $333 in a real month. A numbers-first 30-day bill across four cohorts, with compute, MAU, and egress itemized.
Three modeled 30-day Vercel bills, the $0.15-per-GB meter that actually moves the number, and when Hobby, a VPS, or a bundled host wins.
Three real 30-day Claude Code bills, the 5.5x token-efficiency math, and when the $200 Max plan beats the API.
OpenRouter takes 5.5% on credit purchases with a $0.80 minimum, then passes provider rates through. Here is the real 30-day bill across Anthropic, OpenAI, and Mistral at four workload mixes, with the anti-patterns where the routing tax stops being worth it.
Your billable rate is a sticker price. Your effective hourly rate is the truth. A 2026 cohort breakdown of billable vs effective across five real freelance brackets, plus the math.
A 1,500 customer SaaS shipped AI features and watched gross margin fall from 81% to 47%. We walk three pricing ladders and the COGS math that got it back to 73% in 8 weeks.
Sticker is $20 for both. At a 5-engineer team shipping 100 pull requests a month in June 2026, the all-in 30-day bill diverges. Here is the honest math.
The sticker on Lovable, Bolt, v0, Replit, Cursor, and Totalum is between $20 and $79. The 30-day bill on a real reference workload is $69 to $270. Here is what each pricing page leaves off, with the math.
Three pricing ladders beat flat per-seat pricing for AI products in our teardown of 40 launches: a credit ladder, a usage-with-floor ladder, and an outcome ladder. Each cleared a 7%+ trial-to-paid rate by aligning the price metric with the value metric and putting a visible cap on downside. Pick the ladder that matches how your users feel cost, tokens, runs, or results, and price the rung, not the seat.
The sticker price per token is the smallest line in your LLM bill. Once you add retries, embeddings, vector reads, orchestration, and the platform margin, a "simple" agent workflow lands around $0.42 per run, roughly 4× the raw model cost. This breakdown shows where the money actually goes and which three levers cut a workflow bill the fastest without touching quality.
Build-vs-buy isn't a values debate, it's a breakeven date. SaaS wins early because it converts a big upfront build into a small monthly fee. DIY wins once your usage-based SaaS bill exceeds the fully-loaded cost of owning the code, typically around month 18 for infrastructure-style tools. This piece gives you the formula, a worked example, and the three traps that make teams build too early.
Your headline rate is a vanity number. Your effective hourly rate, total income divided by every hour the business consumed, billable or not, is what you actually earn. For most freelancers it lands 35–55% below the rate on their invoice once you count admin, sales, and unbilled rework. This piece shows how to compute it, why it's always lower than you think, and the three moves that raise it without raising your rate.
