Otaflag — privacy-invisible feature flags
Client-side feature flags and remote configuration without sending
user identities, traits, or targeting data. Apps fetch configuration on
launch and resume, cache it on-device, and evaluate locally. The SDK
collects nothing — no identifiers, no analytics, no tracking — so apps
keep an honest "Data Not Collected" App Store privacy label and require
no ATT prompt.
Guarantees: No user profiles — evaluation happens
on-device. No per-user pricing — plans are based on delivered
configurations. No network round-trip to evaluate a flag. Built to
survive outages — the last known configuration keeps serving. No user
analytics stream. No configuration lock-in — export as JSON and
leave.
Pricing (flat per org per month, metered by
delivered configurations — never seats, never users): Free $0,
~26.25k/mo · Starter $15, ~1.05M/mo · Pro $39, ~5.25M/mo · Scale $99,
~26.25M/mo · Custom from $299. Key packs: +10 credentials/environment
for $5/mo each.
Security model: feature flags are application
configuration, not access control — they must never enforce
permissions, subscriptions, or payments. SDK credentials are read-only
data-plane capabilities designed to ship inside applications; the raw
credential never travels over the network.
Install: npm install @otaflag/sdk
(publishes with public launch). CLI and MCP server for AI agents:
otaflag and @otaflag/mcp.
This site renders in the browser with JavaScript — enable it for
the full site. The complete documentation lives at
/llms.txt.