--- title: "How 30 AI Companies Actually Price Their Products (And What Nobody Tells You About It)" description: "Research-backed data on pricing strategies, business models, and revenue structures for 30 AI companies. Not the polished version they put on their marketing site. The real mechanics of how they charge money." date: 2025-11-28 author: Kat Laszlo canonical: https://tansohq.com/blog/ai-pricing-research --- # How 30 AI Companies Actually Price Their Products (And What Nobody Tells You About It) By [Kat Laszlo](https://www.linkedin.com/in/katrinalaszlo/) ยท November 28, 2025 **Here's something nobody talks about enough:** most AI founders spend weeks agonizing over their pricing model, launch it, then completely rebuild it 6-9 months later. I've watched this happen dozens of times. Smart folks, great products, but they're flying blind because there's no good benchmarking data on how AI companies actually price their stuff. So we built this. Research-backed data on pricing strategies, business models, and revenue structures for 30 AI companies. Not the polished version they put on their marketing site. The real mechanics of how they charge money. --- ## Why This Exists **Most pricing advice is garbage.** You'll read a blog post that says "value-based pricing" or "align with customer success" and think, okay great, but *how* do I actually structure this? Do I charge per API call? Per seat? Per outcome? What's the rollover policy? How do I handle enterprise differently than self-serve? The answer everyone gives: "it depends." That's true, but it's also useless. What actually helps is seeing what your peers are doing. Not to copy them (that's its own trap) but to understand the real trade-offs different models create. To see the details that matter. Like whether Notion AI allows rollover (they don't) or if they have usage caps (they do). These specifics are where pricing actually lives. --- ## What We Documented We analyzed 30 AI companies across the stack. Not a representative sample. We can't claim that. But a useful one. For each company, we documented: **Pricing Mechanics** (the stuff that actually matters): - Rollover policy: Can unused credits carry forward? - Expiration: Do credits expire? - Top-ups: Can you buy more mid-cycle? - Pay-as-you-go: Can you use without commitment? - Usage caps: Are there limits? - Seat pooling: Can a team share an allocation? **Business Strategy:** - Overage strategy: What happens when you hit limits? - Growth strategy: How they acquire customers - Pricing structure: Flat rate, tiered, usage-based? - Revenue model: Subscription, consumption, hybrid? - Value metric: What they actually charge for Take Notion AI as an example. They're bundled into Business/Enterprise plans. Flat rate subscription model. Charge per seat. But where it gets interesting: no rollover, no expiration (because there's nothing to expire), no top-ups, no pay-as-you-go. They have usage caps with a "fair use policy" that'll throttle heavy users. Their overage strategy? "Unlimited with Fair Use." Which really means unlimited until it's not. Heavy users get throttled or contacted. It's honest, but also a bet that most users won't hit the limits. --- ## What We Actually Found After looking at all this data, a few things jumped out: **Most companies don't allow rollover.** Why? Because it creates accounting nightmares and encourages hoarding behavior. You want people using your product, not stockpiling credits "just in case." **"Unlimited" almost always has a catch.** Fair use policies. Throttling during heavy usage. Rate limits buried in the docs. The only true unlimited is when the cost to serve is so low it doesn't matter. **Pay-as-you-go is rarer than you'd think.** Even API companies often require some baseline commitment. Pure PAYG is great for customers but terrible for revenue predictability. Most companies pick a side. **Usage caps show up everywhere in freemium models.** It's the obvious conversion lever. But the companies that nail this set caps that let you do real work, not just kick the tires. Set it too low and people bounce before they see value. **Seat pooling is the hidden feature enterprise customers care about.** Can a 50-person team share a 50-seat allocation flexibly? Or is it locked per person? This seems minor until you're trying to close a deal. --- ## How To Use This **First: don't just copy what someone else does.** I can't stress this enough. Anthropic's pricing works for Anthropic. You're not Anthropic (probably). Instead: **Map out your actual mechanics before you think about price.** Will you allow rollover? What happens at overages? Can people top up? These decisions shape everything else. Notion AI said no to all of it and went flat rate. That's a clear position. **Look at what companies in your category actually do.** Not what they say in blog posts. What they *do*. Check the expiration policies. The rollover rules. The overage strategies. That's where the real decisions live. **Understand the second-order effects.** Allow rollover? Great, now you need to track balances across billing cycles. Fair use policy? Better have a plan for who monitors it and how you communicate with violators. Every choice cascades. **Pick your growth strategy first, then price accordingly.** Product-led growth (like Notion AI) needs generous limits so people can self-serve and experience value. Sales-led needs flexibility to customize. You can't optimize for both. --- ## The Real Challenge **Pricing is a commitment.** Once you launch it, changing it is painful. Your existing customers get grandfathered or angry. Your sales team has to relearn everything. Your billing infrastructure might need a rebuild. That's why most companies get it wrong the first time. You can't really know what the right pricing is until you have customers. But you need pricing to get customers. Classic catch-22. My advice: **pick something, ship it, and plan to iterate.** Just don't iterate every quarter or you'll drive everyone crazy. Give it 6-12 months to see real patterns. And when you do change it, be honest about why. Your customers will respect that more than pretending you had it figured out all along. --- ## Using The Tool Click on any company in the mechanics reference table. You'll see the full breakdown: rollover policy, expiration rules, overage strategy, the works. What to look for: **Companies with similar value metrics to yours.** If you charge per seat, look at other seat-based models. If you charge per API call, look at usage-based. The mechanics that work tend to cluster. **How they handle the edge cases.** What happens when someone hits their limit? How do they think about fairness? These decisions reveal what they actually care about. **The trade-offs they're making.** Notion AI has usage caps but calls it "unlimited." That's a bet that marketing simplicity beats technical accuracy. Is that a bet you want to make? **Growth strategy alignment.** Product-led companies like Notion AI optimize for self-serve conversion. That shapes everything from their pricing structure (flat rate is simple) to their mechanics (no top-ups needed if you're on unlimited). We also built a quick quiz. Answer 4 questions and we'll give you personalized recommendations based on this research. **Is it perfect? No.** Will it solve all your pricing problems? Also no. But it might point you in a useful direction. Or at least help you think through the trade-offs more clearly. --- ## Case Studies We pulled together a few case studies from companies that figured this out: - **The Transparency Shift** (Cursor): What happens when you make pricing dead simple - **The Tier Lock Problem** (Zapier): When your pricing structure becomes a ceiling - **The API Pricing Standard** (Inflection): How usage-based became the default - **The Land and Expand Playbook** (LangSmith): Freemium that actually converts - **The Unlimited Add-On** (Notion AI): When "unlimited" is really a bet These aren't polished marketing stories. They're the messy reality of pricing decisions and their consequences. --- ## One Last Thing **Pricing is product strategy in disguise.** How you charge signals who you're for, what you value, and how you want to grow. Look at Notion AI again. Flat rate subscription. Per seat. Bundled into higher plans. No rollover, no top-ups, no pay-as-you-go. That's not random. That's a company saying: we want predictable revenue, we're selling to teams, and we're optimizing for product-led growth through simplicity. Every pricing decision is really a strategic decision. The mechanics just make it concrete. The companies that get this right aren't smarter than you. They just looked at more data, ran more experiments, and were honest about their trade-offs. That's what this tool is for. The data is here. The patterns are documented. Now the hard part (actually making the call for your business) is still on you. --- ## Want to explore the full data? Click into any company in the mechanics reference. Sort by category, compare across business models, see what companies like yours are actually doing with rollover policies and usage caps and all the details that matter. Or just reach out if you want to talk through your specific situation. I've seen enough of these pricing decisions go sideways that I can probably save you some headaches. *[Kat Laszlo](https://www.linkedin.com/in/katrinalaszlo/) is co-founder of Tanso, flexible pricing infrastructure for SaaS and AI.* ### Want to price with confidence? Book a quick call and we'll walk you through it.