Skip to main content
FinalRun is an open-source, CLI-based tool. You can run it with your own AI provider key (BYOK), or sign up for FinalRun Cloud for managed AI without provisioning a provider account. The questions below cover the most common things people ask.
Only if you want to use FinalRun Cloud for managed AI. Cloud accounts are free to create and include $5 of credits — sign up at cloud.finalrun.app and set FINALRUN_API_KEY. See Cloud API Key.If you’d rather bring your own AI provider key (Google, OpenAI, or Anthropic), no FinalRun account is needed — install the CLI, set the provider key, and run. See AI Providers.
Both run the same CLI and produce the same test artifacts. They differ in how AI access is provisioned and billed:Cloud is the fastest way to get started. BYOK is useful if your team already has a provider account, wants a specific model tier, or needs provider-side billing visibility.
Three AI providers are supported when bringing your own key:Costs are billed directly by your provider at standard API rates. Set the key in your shell, in a .env file at your workspace root, or pass --api-key for a single run.Prefer not to provision a provider account? Use FinalRun Cloud instead — set FINALRUN_API_KEY and you’re done.
  • Android — both real devices (over adb) and emulators are supported.
  • iOS — simulators only today. Physical iOS device support is on the roadmap; join the waitlist for early access.
Android and iOS.
FinalRun the CLI is free and open-source. AI execution is billed in one of two ways:
  • FinalRun Cloud — new accounts include $5 of free credits. After that, runs are billed per execution. Sign up at cloud.finalrun.app.
  • BYOK — you pay your AI provider (Google, OpenAI, or Anthropic) directly for tokens consumed during test execution. Cost depends on the model you choose and how long the test takes. FinalRun adds no markup.
Yes — only on FinalRun Cloud. Set FINALRUN_API_KEY as a CI secret and call finalrun cloud test as a pipeline step:
Local emulator/simulator runs aren’t supported in CI today.
FinalRun ships a set of agent skills that let you generate tests, run them, and fix failures — all from your AI coding agent chat. They work in Claude Code, Cursor, and Codex (and any other agent that supports the skills protocol).The FinalRun installer registers the skills globally. To install or update them manually:
Three slash commands are then available:See AI Agent Skills for details.
Artifacts for each run — including video, screenshots, and device logs — are stored at:
Use finalrun runs to list all recorded runs for your current workspace, and finalrun start-server to open the visual report UI where you can browse results interactively.