Version: v2 · 2026-04-21 · HERALD
Supersedes: BUSINESS-MODEL-PRICING-2026-04-21.md (v1)
Status: advisor-grade, print-ready
Rough-Ask: R0119
> Harnoor: "goal is to use a payment model similar to what Claude uses"
---
Claude's consumer pricing is the cleanest in the AI category. It has four structural properties that match Silent Infinity's ethical stance:
1. Subscription-only, no usage metering shown to the user — the price doesn't change when the user has a hard week.
2. Usage caps, not feature gates — higher tiers raise ceilings; they don't lock features behind paywalls.
3. Flat monthly rate + discounted annual — no dynamic pricing, no "save 80% today" discount theater.
4. No ad tier, no data sale, no affiliate referral kickbacks — revenue comes from subscribers, full stop.
This is structurally identical to what the Emergent Constellation plan and Brand Book already commit to. Adopting Claude's model is less a strategic choice than an alignment correction — v1 tried to engineer value via tier-specific features (voice at Pro, clinical referral at Pro, HIPAA at Org), which created the math flags v2 fixes:
V2 solves all three by raising ASP to $20 (Claude Pro parity) and dropping features that require external vendor relationships.
---
The free tier is not a funnel to conversion. It is the product's ethical floor: anyone should be able to use Silent Infinity in a moment that matters without a credit card.
---
Pricing Silent Infinity Plus at $20 places it in the "AI tool" bucket, not the "meditation content app" bucket. That's the correct mental frame for buyers: they are paying for an ongoing reflective partner, not a library of pre-recorded meditations.
The $9 price point in v1 signaled "cheap meditation app" which compresses long-term willingness to pay. Raising to $20 matches where the product actually sits.
$100 captures the power-user segment. Opus 4.7 routing alone is worth the delta for users who want the deepest reflection.
Slightly more generous than Claude because early-stage cash cost is meaningful.
---
Assumptions (documented for audit):
| Tier | Blended monthly revenue per user | LLM cost | Payment processor (2.9 % + 30¢) | Gross profit | Gross margin |
|---|---|---|---|---|---|
| Free | $0 | -$1.80 (12 turns avg × $0.0075 × 20 days) | $0 | -$1.80 | (marketing cost) |
| Plus | $17.50 | -$2.25 | -$0.81 | $14.44 | 82 % |
| Max | $85 (monthly blend) | -$11.25 (Opus 4.7 on every turn × 30/day) | -$2.77 | $70.98 | 84 % |
| Team | $28 / seat | -$2.25 | -$1.13 | $24.62 | 88 % |
| Enterprise | custom (avg $150 / seat) | -$3.00 | negotiable | ~$140 | ~93 % |
All tiers comfortably above 80 % gross margin. The v1 Pro-at-22%-margin problem is gone because we dropped the counselor-referral commitment.
Free tier economics: -$1.80 / user / month fully-loaded. Funded as marketing from the subscriber book. Break-even requires 0.13 paid users per 10 free — well within consumer conversion norms (typically 3-8 %).
---
| Metric | Year 1 | Year 2 | Year 3 |
|---|---|---|---|
| DAU | 2,500 | 40,000 | 250,000 |
| Free : Plus : Max : Team conversion | 95 : 4 : 0.5 : 0.5 | 93 : 5 : 1.0 : 1.0 | 90 : 7 : 1.5 : 1.5 |
| Plus subscribers | 100 | 2,000 | 17,500 |
| Max subscribers | 12 | 400 | 3,750 |
| Team seats | 12 | 400 | 3,750 |
| Monthly revenue (Plus) | $1,750 | $35,000 | $306,250 |
| Monthly revenue (Max) | $1,020 | $34,000 | $318,750 |
| Monthly revenue (Team) | $336 | $11,200 | $105,000 |
| Total MRR | $3,106 | $80,200 | $730,000 |
| ARR | $37k | $963k | $8.76M |
Year 3 ARR of $8.76M is within 10 % of v1's $9.7M target — but this time with defensible math. The conversion rates are conservative; 5 % Plus is below the 7-10 % ChatGPT achieves and the 4 % Duolingo Premium hits. We can beat these if the product is good.
---
Maintaining the ethical stance requires explicit NO on:
1. Advertising — no banners, no sponsored content, no affiliate links in responses
2. Data sale — raw or anonymized user data never sold
3. Training-data licensing for third-party models — only our own fine-tuned derivatives may be licensed, under IRB review
4. Usage-based upsell — we will never charge $0.10/extra turn. Caps soft-fail with dignity ("try again in N hours" + the user is not shamed)
5. Surge pricing — no dynamic personalization based on user wealth signals
6. Attention capture SKUs — no streaks-for-pay, no "skip waiting" packs, no consumable micro-payments
7. Affiliate referral commissions on therapy — cannot have undisclosed incentives to refer out
8. Medical claims — Plus/Max remain wellness tiers, not medical. No "therapeutic" marketing copy
---
1. Grandfathering — current subscribers lock their price on annual plans, forever
2. Donation tier (Innovation 6) — a 501(c)(3) structure holds the crisis-detection layer as a public good; donations accepted against tax-deductible receipts. Never required; never nudged at the user
3. Student + low-income rate — $5 / mo Plus, self-selected via honor system, no documentation required
4. Research partnership grants — academic IRB-approved studies can pay us for anonymized aggregate metrics at cost, not profit
5. Wisdom library royalties — if we commission original works (poetry, essays, short audiobooks) from contemporary contemplative writers, revenue share is direct to them
6. Enterprise contracts — full negotiation latitude; the only tier where price is not published
---
---
| Item | v1 | v2 | Reason |
|---|---|---|---|
| Plus price | $9 / mo | $20 / mo | Claude-parity; v1 priced as meditation app, not AI tool |
| Pro tier | $19 / mo | Removed | Created margin exposure via counselor referral |
| Max tier | Not present | $100 / mo | Captures power users; Claude-parity |
| Organization | $99 / seat | Split into Team $30 / Enterprise custom | Clearer value ladder |
| Feature gating | Voice at Pro, HIPAA at Org | All features on Plus; caps differentiate tiers | Matches Claude philosophy |
| Counselor referral | $5 / mo cost at Pro | Dropped | Eliminates unsigned-vendor exposure |
| Year 3 ARR projection | $9.7M | $8.76M | Defensible unit economics |
---
No price increase for grandfathered users without 12-month written notice. No "pay to skip dark pattern" — we have no dark patterns. No dynamic / personalized pricing. Annual transparency publishing of pricing decisions + margin data in the quarterly transparency report.
Every pricing change documented with: proposed change, rationale, projected impact, customer-impact analysis, 30-day public comment period, final decision, implementation date.
---
— HERALD · 2026-04-21