Starter stack

AI app that bills by usage

Every request costs you model tokens, so a flat subscription either loses money on heavy users or prices out light ones. You want to charge per use, credits or tiers with limits — and see what each user costs you.

8 makers' products and 10 open-source projects use at least half of these 9 tools.

What it costs

Launchabout 1,000 users: $1.74/mo
Tractionabout 10,000 users: $98/mo
Scaleabout 100,000 users: $1,542/mo

Plus fees on revenue (Stripe): $44 / $410 / $4,100 at those sizes.Monthly bills of the priced picks, list prices.

Do the picks fit together?No conflicts: no acquired or archived tool, no two tools for a decision that usually has one, nothing the plan's rules say doesn't fit.Open in Check my stack — add or swap tools →

Hand it to your AI agent

Paste into AGENTS.md, CLAUDE.md or your Cursor rules: each tool with its agent docs, MCP server and CLI where they exist.

Preview

# Stack: AI app that bills by usage

Use these tools. Ask before swapping one for an alternative.

- Frontend Framework: Next.js — https://nextjs.org
agent docs: https://nextjs.org/llms.txt · CLI: npx next
Why: Streaming responses and API routes in one app, where usage gets metered.
- Authentication: Clerk — https://clerk.com
agent docs: https://clerk.com/llms.txt · MCP server: io.github.clerk/mcp-server (https://mcp.clerk.com/mcp) · CLI: npx clerk
Why: Prebuilt sign-in and organizations, so usage can be tracked per user or per team from the start.
- Database: Convex — https://convex.dev
agent docs: https://convex.dev/llms.txt · CLI: npx convex
Why: A realtime TypeScript backend, so credit balances and usage update on screen the moment a request finishes.
- LLM API: OpenRouter — https://openrouter.ai
agent docs: https://openrouter.ai/llms.txt
Why: One key and one bill for hundreds of models, so you can route cheap requests to cheap models and keep margins.
- AI SDK & Agent Framework: AI SDK — https://ai-sdk.dev
Why: Returns token usage with every call, which is what you meter and bill.
- Payments: Autumn — https://useautumn.com
agent docs: https://useautumn.com/llms.txt · CLI: npx atmn
Why: Credits, limits and usage-based plans on top of Stripe, without writing the metering and entitlement logic yourself.
- Payments: Stripe — https://stripe.com
agent docs: https://stripe.com/llms.txt · MCP server: com.stripe/mcp (https://mcp.stripe.com) · CLI: npx @stripe/cli
Why: The processor under the billing layer, and the one most makers here already use.
- LLM Observability & Evals: Langfuse — https://langfuse.com
agent docs: https://langfuse.com/llms.txt · CLI: npx langfuse-cli
Why: Cost and tokens per user and per request — what you need to price usage — open source, free to self-host or on its hosted plan.
- Hosting: Vercel — https://vercel.com
agent docs: https://vercel.com/llms.txt · MCP server: com.vercel/vercel-mcp (https://mcp.vercel.com) · CLI: npx vercel
Why: Serverless functions that scale with request volume, matched to a bill that grows with usage too.

Trade-offs: Usage pricing is harder to explain than a flat plan, and users worry about surprise bills — show live usage and let them set limits. Several pieces (billing layer, gateway, observability) each take a small cut or a monthly fee; the simplest start is Stripe's own usage billing, adding the rest when pricing gets complex.

Source: https://howmakersbuild.com/stacks/ai-usage-billing

The picks

Frontend Framework

Next.jsllms.txt CLI

Streaming responses and API routes in one app, where usage gets metered.

Chosen for frontend framework by 1506 makers' products we track · the alternatives →

Authentication

Clerkllms.txt MCP CLIIn Replit

Prebuilt sign-in and organizations, so usage can be tracked per user or per team from the start.

Chosen for authentication by 305 makers' products we track · the alternatives →

Database

Convexllms.txt CLI

A realtime TypeScript backend, so credit balances and usage update on screen the moment a request finishes.

Chosen for database by 50 makers' products we track · the alternatives →

LLM API

OpenRouterllms.txtIn Replit

One key and one bill for hundreds of models, so you can route cheap requests to cheap models and keep margins.

Chosen for LLM API by 61 makers' products we track · the alternatives →

AI SDK & Agent Framework

AI SDK

Returns token usage with every call, which is what you meter and bill.

Chosen for AI SDK & agent framework by 4 makers' products we track · the alternatives →

Payments

Autumnllms.txt CLI

Credits, limits and usage-based plans on top of Stripe, without writing the metering and entitlement logic yourself.

Chosen for payments by 12 makers' products we track · the alternatives →

Payments

Stripellms.txt MCP CLIIn Bolt, Lovable, Replit, v0

The processor under the billing layer, and the one most makers here already use.

Chosen for payments by 869 makers' products we track · the alternatives →

LLM Observability & Evals

Langfusellms.txt CLI

Cost and tokens per user and per request — what you need to price usage — open source, free to self-host or on its hosted plan.

Chosen for LLM observability & evals by 33 makers' products we track · the alternatives →

Hosting

Vercelllms.txt MCP CLIIn v0

Serverless functions that scale with request volume, matched to a bill that grows with usage too.

No maker's product we track shows it for hosting yet · the alternatives →

What this costs you

Usage pricing is harder to explain than a flat plan, and users worry about surprise bills — show live usage and let them set limits. Several pieces (billing layer, gateway, observability) each take a small cut or a monthly fee; the simplest start is Stripe's own usage billing, adding the rest when pricing gets complex.

Ask your AI about this, with this page as the source:ChatGPT ↗Claude ↗Perplexity ↗

Who builds this way

Using at least half of these tools, each with a source: makers' products from their own words, customer stories, subprocessor lists or websites; open-source projects from their code.