Launching soon — join the private beta

Know which customers make you money — and which burn it

ProfitLens joins your Stripe revenue with your real LLM token costs and shows a live margin for every customer. No guesswork, no “AI insights” — just your own numbers, multiplied correctly.

Founding users get the beta free · One email at launch, no spam

2,200+ models priced live Stripe-native revenue No code to start
ProfitLens · Dashboard

Gross margin

34.2%

Revenue (MRR)

$1,693

AI cost

$676.43

Revenue vs AI cost — 2026-07, by day

$60$45$30$15$0
0104071013161922
AI CostRevenue
CustomerMRRAI costMargin
LegalCorp$499$182.40+$316.60
Nova Firm$299$96.15+$202.85
Casewise$299$121.02+$177.98
Davis Law$199$235.09−$36.09
Atlas Legal$99$41.77+$57.23

Cost anomaly detected

Davis Law spent 5.2× its daily average yesterday

Unit economics is the blind spot of AI products

Every AI SaaS has customers that cost more than they pay. Most teams find out at the end of the quarter — from the provider invoice.

Flat-fee pricing, metered costs

You charge $199/mo, but tokens are billed per request. One power user on a heavy model quietly turns a good account into a loss.

Revenue and costs live apart

Stripe knows who pays you. Provider consoles know what you spend — but only in aggregate. Nobody joins the two per customer.

Prices change under your feet

Providers reprice models mid-quarter, caching changes the math, new models ship monthly. Spreadsheets can't keep up.

Customer pays

$199/mo

Their token spend

$235.09

=

Your margin

−$36.09

…and this customer looks perfectly healthy in Stripe. You'd never know.

Everything between “we spend a lot on tokens” and “we know exactly where”

Six tools, one goal: know your real margin at any moment — and fix it before the invoice does.

Per-customer margin

Every customer gets a live P&L: MRR from Stripe minus their actual token spend, by month. Losing accounts turn red.

One-click Stripe connect

OAuth in, customers and MRR sync automatically. Webhooks keep the dashboard current when plans change.

Real model prices, always fresh

Token prices for 2,200+ models synced from the LiteLLM catalog. Price changes are versioned by date — history stays exact.

Loss & anomaly alerts

Get pinged when a customer goes margin-negative or spends multiples of their daily average — before the invoice does.

Repricing simulator

Drag a plan price and instantly see the new gross margin and which customers flip back to profitable. Apply in one click.

Import any usage export

Drop in the CSVs from the OpenAI or Anthropic console — messy headers, totals rows and duplicates handled automatically.

From zero to unit economics in one afternoon

01

Connect Stripe

One OAuth click pulls your customers, plans and MRR. No keys to copy, nothing to export.

02

Add your token usage

Import provider usage CSVs to start, or stream events from your backend — one field of integration: the customer id.

03

See margins. Act.

The dashboard ranks customers by margin, alerts flag the bleeders, and the simulator shows the price that fixes it.

~20 minutes to your first margin view1 field to integrate — the customer idRead-only Stripe access, revocable anytime

Costs are computed from live token prices — 2,200+ models, every major provider, always current

GPT-5OpenAIGPT-5 MiniOpenAIGPT-4oOpenAIo3OpenAIClaude Opus 4.8AnthropicClaude Sonnet 4.6AnthropicClaude Haiku 4.5AnthropicGemini 2.5 ProGoogleGemini 2.5 FlashGoogleGrok 4xAIDeepSeek V3DeepSeekMistral LargeMistralLlama 4 MaverickMetaNova ProAmazon

Be first to see your real margins

We're opening the private beta to a small group of AI SaaS teams. Founding users get it free during the beta — and keep a discount for life.

Free during the betaLifetime founder discountDirect line to the founder

One email when we launch. No newsletters, no spam.