How you charge for your app shapes every other decision — which features ship, which users you attract, how you grow, how much support burden you take on. Picking the wrong model is more expensive than picking a slightly-wrong price within the right model. This post is the 2026 decision guide.
The six models
- Free — no monetization, app is free with no purchases or ads.
- Paid up-front — one-time purchase to download.
- Freemium — free download with a paid unlock or premium tier.
- Subscription — recurring monthly or annual fee.
- IAP / consumables / credits — buy units of usage.
- Ad-driven — free app, revenue from displayed ads.
Plus hybrids that combine these (covered below).
Free (no monetization)
Use when: the app is a marketing channel for another product, a portfolio piece, a learning project, or a community service. Don't use when: you intend to support the app long-term and need it to pay for itself.
Reality check: "free for now, add monetization later" is the way to never make money. Plan monetization at launch or after a clear validation milestone.
Paid up-front
One-time price ($0.99-$9.99 typical for indie; $19.99+ for prosumer/pro tools). User pays, downloads, owns forever.
Works for:
- Premium utilities (high-quality calculators, photo tools, productivity apps).
- Games where the value is in the content, not ongoing service.
- Apps with no per-user variable cost (no AI / server bills).
- Audiences that distrust subscriptions (creative pro tools, certain dev tools).
Fails when: there's significant ongoing cost (AI calls, backend scale, content updates) — revenue stops at install but costs continue. This is why RDR2 Companion's $2.99 one-time unlock is tight margin when AI usage continues.
App Store reality: paid-up-front means fewer downloads. People test free apps freely; they hesitate on paid. Conversion to install is 10-20x lower than free.
Freemium (the dominant indie pattern)
Free app, paid unlock for premium features or limit removal. RDR2 Companion's $2.99 unlock after 4 free AI questions is exactly this pattern.
Why it works:
- Low download friction — anyone can try.
- Value demonstrated before payment.
- Clear "moment of value" triggers the upgrade decision.
- Refund rate is lower than paid up-front (users already saw what they're buying).
Critical design choices:
- The trial / free tier must show real value — not a crippled demo. 4-10 meaningful uses, or 7-30 days of real access.
- The paywall moment must feel earned, not blocked. Show it when the user has just experienced value and would want more.
- Conversion rate of 3-7% is healthy for consumer freemium; 1-3% is typical; below 1% means the value isn't clear before the gate.
Subscription
Monthly or annual recurring fee. Dominant model for productivity, fitness, AI, media, dating.
Use when:
- Your app has ongoing variable cost (AI calls, server, content updates).
- Value is delivered continuously (productivity, fitness, news, AI assistant).
- You can credibly add value over time (new features justify continued payment).
- The audience accepts subscriptions for your category (consumers will subscribe to fitness, productivity, AI; they resent it for utilities, calculators, simple games).
Key knobs:
- Monthly vs annual. Annual at 30-50% discount: lower churn, front-loaded cash. Most subscription apps offer both; many users pick annual after a monthly trial.
- Free trial. 3 days, 7 days, 30 days are standard. Longer trials reduce conversion (users forget) and increase abuse. 7 days is the sweet spot.
- Price points. $4.99/mo or $39.99/yr is a common indie tier. $9.99/mo or $79.99/yr for premium tools. Test ceiling — usually higher than instinct.
Apple Small Business Program brings the 30% to 15% on first-year subs (and all subsequent). For an indie this is meaningful — apply.
IAP / Consumables / Credits
Buy units of usage. Especially well-fit for AI apps where cost scales per call.
Pattern: users buy "100 AI questions" for $4.99, "500 for $19.99", etc. Each question costs an actual amount you can predict.
Why it's good for AI apps: aligns revenue with cost. Heavy users pay more (which they should because they cost more). Light users pay less. No surprise unprofitable users.
Tradeoffs: some users find it nickel-and-diming. Mixed perception in the App Store. Reserve for genuine usage-based products, not as a tax on every interaction.
Ad-driven
Free app with displayed ads. Revenue per user is low.
Use when:
- App has no per-user variable cost (no AI bills).
- Audience won't pay anything (younger demographics, frequent-use entertainment).
- Engagement is high enough that ad impressions add up.
Don't use when:
- Your app is AI-powered — ad revenue rarely covers AI cost.
- Your audience is professional / B2B — they expect paid quality.
- You want users to actively pay attention — ads degrade attention.
Typical eCPM in 2026: $1-$10 depending on demographic and ad format. A user with 20 monthly impressions earns the app $0.02-$0.20/mo. You need millions of users for ads to make real money.
Hybrid models
Most successful 2026 apps mix:
- Freemium + subscription tier. Free with a limit, $4.99/mo to remove it. Most flexible.
- Freemium + one-time + subscription. Free tier, $2.99 unlock for basic, $9.99/yr Pro for power features. RDR2 could evolve here.
- Subscription + IAP credits. Subscribers get a monthly credit allowance, can buy more.
- Free + ads + IAP to remove ads. Casual games standard.
- Free + paid IAP for cosmetics / boosts. Mobile games standard.
Decision tree
- Does the app have per-user variable cost? (AI, backend that scales, content updates)
- YES → skip pure paid up-front. Subscription, freemium with sub tier, or IAP credits.
- NO → paid up-front or freemium both work.
- Is the value continuous (ongoing) or one-time?
- Continuous → subscription wins.
- One-time → paid or freemium-unlock.
- Does the audience accept subscriptions in your category?
- YES → subscription is the highest LTV option.
- NO → freemium with one-time unlock is friendlier.
- How large is your potential audience?
- Small (<10K likely buyers) → higher prices, subscriptions, B2B angle.
- Large (millions) → can monetize at low ARPU with freemium / ads.
- What's your appetite for support load?
- Subscriptions have higher support burden (refund requests, billing disputes).
- One-time paid is the lightest.
AI app special considerations
AI apps in 2026 have unique economics that shape monetization:
- Per-user cost is real and ongoing. Pure paid up-front doesn't work long-term; you need recurring revenue or usage-based pricing.
- Heavy users cost real money. Either rate-limit free tiers aggressively, or use credit-based pricing to align cost and revenue.
- Model swap is a margin lever. Cheap routine queries on Haiku, expensive ones on Sonnet/Opus — visible in your bill but invisible to users.
- Prompt caching is the highest-leverage cost reduction. 60-90% input cost reduction on cached system prompts. Implement this.
- Don't promise "unlimited" AI without rate limits. A power user can run up $100/mo in your costs while paying $9.99.
For RDR2 Companion specifically: the v1.2 freemium $2.99 unlock works for casual users but should add a Pro tier (perhaps $9.99/yr) for power users who would otherwise be unprofitable. Use the profitability calculator to model the threshold.
See: Framework, iOS Economics, Profitability Calculator, Marketing Fundamentals.