For platforms
For AI app builders, white-label app platforms, and template studios: a hosted API that writes and localizes your users' store listings, then stages them for publishing through each connected console — the part of shipping an app your product doesn't cover and humans don't scale to.
REST + OpenAPI 3.1 + MCP · Per-tenant credentials · Charged only on success
curl -X POST "https://appdrift-backend-1fabfc95f592.herokuapp.com/v1/generate-metadata" \
-H "Authorization: Bearer ad_live_..." \
-H "Content-Type: application/json" \
-d '{
"platform": "ios",
"app_name": "Drift",
"app_description": "A minimalist habit tracker
with streaks and smart reminders.",
"fields": ["subtitle", "keywords"],
"language": "en-US"
}'The loop
The listing lifecycle as four API surfaces — shapes below are the real ones from the API reference, not marketing pseudocode. Publishing itself runs through each tenant's connected console, so nothing reaches a live store un-reviewed.
POST /v1/generate-metadataASO titles, subtitles, descriptions, and keyword fields written from a short brief — 1–2 tokens per field, charged only on success.
POST /v1/translateAny metadata field into up to 15 languages per call, informed by live per-market store data — 1–5 tokens per language.
POST /v1/apps/{id}/optimize/applyAPI writes land in a reviewable draft; publishing to App Store Connect and Google Play runs through AppDrift's connected-console flow with each tenant's own credentials — nothing reaches production un-reviewed.
GET /v1/apps/{id}/keywords + webhooksDaily ranks with difficulty scores per tracked keyword, and webhooks (keyword.rank_change, monitoring.alert, action_plan.ready) that push changes into your stack.
curl -X POST "https://appdrift-backend-1fabfc95f592.herokuapp.com/v1/generate-metadata" \
-H "Authorization: Bearer ad_live_..." \
-H "Content-Type: application/json" \
-d '{
"platform": "ios",
"app_name": "Drift",
"app_description": "A minimalist habit tracker
with streaks and smart reminders.",
"fields": ["subtitle", "keywords"],
"language": "en-US"
}'{
"error": null,
"message": "Success",
"data": {
"platform": "ios",
"language": "en-US",
"fields": {
"subtitle": "Build habits that stick",
"keywords": "habit,tracker,routine,streak,daily,goals"
},
"tokens_charged": 2
}
}Full endpoint list, auth, webhooks, and error envelope in the API reference; machine-readable OpenAPI 3.1 at /.well-known/openapi.json on the API host. Agent workflows go through the MCP server.
Why this exists
Both of today's options are real — they just weren't built for a platform shipping other people's apps.
Free and official — and a grind: JWT signing against App Store Connect, service-account plumbing for Play, per-account quirks, and zero content layer. They upload whatever text a human already wrote. If nobody wrote it, there's nothing to upload.
Excellent for a human managing one portfolio — but they sell seats and screens, not an API your product can integrate. A platform shipping fifty client apps a month can't put a human in fifty dashboards.
Nobody sells the loop — write → localize → stage → track — as a hosted API. That's the gap this fills.
And the tenancy model fits the way platforms already operate: Apple's guideline 4.2.6 pushes template and white-label apps onto per-client developer accounts, and AppDrift runs per-tenant App Store Connect and Play credentials — one tenant per client, cleanly isolated.
What we don't do (yet)
Pricing
Usage is metered on the same tokens as the rest of AppDrift and charged only on success — a failed call costs nothing, which makes a pilot hard to lose money on. Platform volumes vary too much for one public rate card to be honest, so bring your shape — apps per month, locales per app, publish cadence — and we'll quote it straight.
Tell us your shapeFAQ
With per-tenant API keys. Every workspace mints its own bearer key (ad_live_…) that only ever sees that tenant's apps and credentials — shown once, stored hashed, revocable any time. Store access works the same way: each tenant connects its own App Store Connect API key and Google Play service account through AppDrift's standard connect flow. There is no shared super-key and no cross-tenant visibility.
Yes — that's the model. Apple's guideline 4.2.6 pushes template and white-label platforms toward publishing each client's app from the client's own developer account, and AppDrift's per-tenant App Store Connect and Play credentials map onto that one-to-one: one tenant per client, with that client's own store connection inside it. Today each tenant is created through the normal signup and connect flow and mints its own API key, which your platform then drives. Programmatic tenant provisioning — creating workspaces and attaching store credentials over the API — is the piece we build with a launch partner rather than ahead of one; tell us your onboarding shape and we'll scope it.
Volume pricing, by conversation. Usage is metered on the same tokens as the rest of the product and only charged on success — a failed call never costs anything. We don't publish a platform rate card yet because volumes vary wildly between a template studio and an AI app builder; tell us your shape (apps per month, locales, publish cadence) and we'll quote it straight.
The API is plain REST with a machine-readable OpenAPI 3.1 spec published at /.well-known/openapi.json on the API host, so any language with an HTTP client works today — and you can generate a typed client from the spec. There's also a first-class MCP server, which means AI agents (Claude, Cursor, or your own) can drive the same capabilities directly. Docs cover both.
Tell us what your users do after export today. If the API isn't the right fit, we'll say so — the conversation is the point.
Every platform that ships apps for its customers — AI app builders generating React Native and Flutter projects, white-label builders publishing client apps, template studios selling launch kits — hits the same wall: the binary ships, and then the listing work starts. Metadata, keyword fields, descriptions, localization, publishing to two consoles. A platform charging $50 a month per customer can't spend $500 of human time per listing, so the work lands on end users, and their apps launch with empty, unlocalized, unoptimized listings. The Store-Ops API turns that layer into infrastructure: REST endpoints that generate metadata, localize it into 40+ languages using live per-market store data, publish through AppDrift's connected-console flow, and track the resulting ranks with webhooks.
The same surface is exposed two ways: plain REST with an OpenAPI 3.1 spec you can generate clients from, and a first-class MCP server so AI agents can drive listing operations directly. Write endpoints are deliberately conservative — AI output lands in a reviewable draft rather than going straight to production, and every AI call is charged only on success. Rate limits are per key; each tenant's App Store Connect and Google Play credentials stay isolated to that tenant.
This page is for companies whose product ships other people's apps. If you're a developer optimizing your own app, you want the AppDrift platform overview — same engine, human-sized. And if you're a single team that just wants your listing done for you once, the fixed-price services exist for exactly that.