Sign up for free, upload your APK or enter your web URL, and describe your first test in plain English. No SDK, no code changes, no setup required. You can run your first test in under 60 seconds.
No. Doksi tests your app externally using AI vision—no SDK, no instrumentation, no code changes. Just upload your APK or point us to your web app.
We support Android (APK uploads), Web, and iOS (coming soon). Tests run on cloud-based emulators for speed. Real device testing is on our roadmap.
Yes. Doksi is free to start with no credit card required. You get full access to all features.
Traditional tools require brittle, selector-based scripts that break whenever your UI changes. Doksi uses AI vision models to understand your app like a human would. You describe tests in plain English, and Doksi figures out how to execute them—no XPath, no CSS selectors, no maintenance.
Doksi uses vision models trained on millions of UI samples. It understands intent ("tap the login button") rather than relying on specific element IDs or XPath. This means tests don't break when your UI changes.
On the first run, Doksi explores your app and learns its structure. This takes 10-15 minutes. Every subsequent run uses cached intelligence and completes in under 3 minutes.
When UI elements change—renamed buttons, new layouts, restructured flows—Doksi recognizes the intent and adapts automatically. If "Login" becomes "Sign In", Doksi finds it. No test maintenance required.
You provide context—PRDs, specs, or just a description of what you want to test. Doksi analyzes this to understand your app's key flows and creates tests accordingly.
Functional bugs (broken flows, unresponsive buttons), API issues (backend errors, timeouts), UI/UX issues (missing elements, layout problems), platform-specific bugs, and edge cases like OTP flows and OAuth.
Yes. Doksi intercepts SMS and email OTPs automatically—no manual input needed. This is one of the things that makes traditional automation fail.
Handled. Doksi navigates WebViews, OAuth popups, social logins, and embedded browsers like a real user would.
Yes. Doksi can switch between apps, handle share sheets, deep links, and system dialogs. Test flows like "share to WhatsApp" or "open link in browser and return to app."
Yes. Test your app on 3G, offline, or unstable connections to catch edge cases before your users do.
Everything you need to debug: screenshots at every step, full video replay, network logs with request/response data, device metrics (CPU, memory), and a plain-English summary of the root cause.
Yes. Doksi captures all HTTP traffic including status codes, response times, and response bodies. When a test fails because an API returned a 500, you'll know immediately.
Every test run generates a shareable link. Anyone with the link can view the full results—screenshots, video, logs—no login required.
Yes. Click "Create Issue" on any failure to generate a ticket with all the context (screenshots, logs, root cause) included.
Multiple ways: Slack (@Doksi run checkout flow), GitHub Actions (on every PR), CI/CD webhooks (Jenkins, CircleCI, GitLab), scheduled runs (nightly regression), or manual one-click from the dashboard.
Yes. Connect your repo and Doksi can run tests automatically on every PR, pull context from your codebase, and post results as PR comments.
Full Slack integration: trigger tests with @Doksi commands, get instant notifications on failures, receive daily digest reports, and share results directly in channels.
You can create issues directly from failed tests. Native integrations with Jira, Linear, and GitHub Issues are on the roadmap.
Yes. The free tier includes full functionality with no credit card required. We're focused on getting feedback and building the best product.
One execution of one test on one platform. Running the same test on iOS, Android, and Web counts as 3 runs.
Yes, eventually. We'll introduce paid tiers as we add more capabilities (real devices, higher concurrency, enterprise features). Early users will be grandfathered into favorable pricing.
Yes. Tests run in isolated, ephemeral environments. No app data is stored after test completion. All connections are encrypted. We don't access your backend systems.
APKs are stored only for the duration needed to run tests. You can delete them anytime from your dashboard.
Doksi only interacts with what you provide—test accounts on staging environments, demo accounts, or whatever you configure. We recommend using test environments rather than production.
QA Wolf is a fully managed service where their team writes and maintains tests for you (~$90K/year). Doksi is self-serve—you describe tests in plain English, and AI handles execution and maintenance. Much more affordable, much faster to start.
Mabl uses record-and-playback with some AI assistance. Doksi is AI-native—no recording, no selectors, just describe what you want to test. Mabl also has opaque enterprise pricing; Doksi is free to start.
Maestro uses YAML-based test definitions—simpler than code, but still requires writing test scripts. Doksi uses plain English. Maestro is also primarily a framework you self-host; Doksi is fully managed cloud infrastructure.
Discord for quick questions and community discussions, email support@doksi.ai for direct support, or use the in-app chat widget for immediate assistance.
Absolutely. We're building Doksi based on user feedback. Drop feature requests in Discord or email us directly. We ship fast.
Yes. Visit docs.doksi.ai for guides, API reference, and tutorials.
Can't find the answer you're looking for? Reach out to our team.