SaaS teams shipping AI to their own users

Embed paid AI features in your SaaS

No subscription, pay only for inference
Use with any model
No backend required
Per-customer limits and analytics
Hide system prompts
Tool use and structured output included

Start free See SDK

Flow demo

sdk + webhook

Track one customer

browser-safe

import { createApiKeySdk } from "keystash-sdk";

const sdk = createApiKeySdk({
  baseUrl: "https://api.keystash.dev"
});

await sdk.account.connect(process.env.KEYSTASH_APP_KEY!);

const session = await sdk.runtime.sessions.create({
  spaceId: "acme-prod",
  endUserId: "user_123",
  operationIds: ["summarize_ticket"],
  ttlMinutes: 15
});

await sdk.runtime.connect(session.token);

Result

{
  "spaceId": "acme-prod",
  "endUserId": "user_123",
  "operationIds": ["summarize_ticket"],
  "expiresIn": "15m",
  "token": "kst_rt_..."
}

One customer id per session
Per-customer analytics start here
Prompts and provider keys stay off the client

Point Stripe to Keystash

space config

const sdk = createApiKeySdk({
  baseUrl: "https://api.keystash.dev"
});

await sdk.account.connect(process.env.KEYSTASH_APP_KEY!);

const billing = await sdk.spaces.endUserBilling.update("acme-prod", {
  stripeWebhookSecret: process.env.STRIPE_WEBHOOK_SECRET!,
  endUserIdSource: "metadata",
  endUserIdMetadataKey: "keystash_end_user_id",
  spendAmountSource: "metadata",
  spendAmountMetadataKey: "keystash_spend_usd"
});

console.log(billing.webhookUrl);

Result

{
  "webhookUrl": "https://api.keystash.dev/api/spaces/acme-prod/end-user-billing/stripe/webhook",
  "endUserIdSource": "metadata",
  "spendAmountSource": "metadata"
}

Stripe talks to Keystash directly
No owner backend required for spend grants
One space can meter many end users

Invoke one feature

runtime

const result = await sdk.runtime.invoke({
  operationId: "summarize_ticket",
  input: {
    ticket: "Customer asks for refund after duplicate charge."
  }
}, session.token);

Result

{
  "operationSlug": "summarize_ticket",
  "model": "openai/gpt-5.4-mini",
  "outputText": "Duplicate charge. Refund likely. Escalate billing.",
  "usage": {
    "inputTokens": 212,
    "outputTokens": 36,
    "totalTokens": 248
  }
}

Usage is tracked per customer and per feature
Remaining spend is deducted from that customer
Output returns through one runtime API

Stripe completes a payment

direct webhook

{
  "type": "checkout.session.completed",
  "data": {
    "object": {
      "metadata": {
        "keystash_end_user_id": "user_123",
        "keystash_spend_usd": "25"
      }
    }
  }
}

Result

{
  "endUserId": "user_123",
  "grantedSpendUsd": 25,
  "source": "stripe.checkout.session.completed"
}

Spend allowance is attached to that end user
Runtime can block when the user runs out
Owner keeps Stripe and product UI on the client

Customer

Scoped access

One runtime session per customer and feature set.

Billing

User spend caps

Stripe payments can increase one user's remaining spend.

Prompts

Hidden by default

Ship UI to the browser without exposing prompt logic.

Define

One feature per product action
Provider key stays server-side
Prompt stays hidden

Run

Mint runtime sessions per end user
Invoke features from browser or app shell
Keep models and prompts scoped

Meter

Track calls and token usage by workspace
Track usage by feature
Debit owner balance as users run features

SDK first

`sdk.spaces.operations.create(...)`
`sdk.runtime.sessions.create(...)`
`sdk.runtime.invoke(...)`
`sdk.runtime.usage(...)`

No backend required

Ship a client-side app with one app key. KeyStash mints short-lived runtime sessions for each customer.

Hidden system prompts

Feature prompts live in KeyStash. The browser only sends input and receives output.

Optional backend

If you already run your own checkout flow, Stripe can still call Keystash directly to increase one end user's remaining spend.

Create feature

Admin

const feature = await sdk.spaces.operations.create("acme-prod", {
  title: "Summarize ticket",
  slug: "summarize_ticket",
  model: "openai/gpt-5.4-mini",
  systemPrompt: "You summarize support tickets for internal agents.",
  userTemplate: "Ticket:\n{{ticket}}\n\nReturn a concise summary.",
  monthlyInvocationLimit: 10000,
  allowedOrigins: ["https://app.acme.com"]
});

Dashboard preview

summarize_ticket

openai/gpt-5.4-mini

active

Calls

12,842

Tokens

3.2M

Runtime sessions

TTL 15 min

Origin lock app.acme.com

What ships now

Any-model routing
Client-side app flow
Per-customer limits
Per-customer analytics
Hidden system prompts
Tool use and structured output