Dimension scores are derived from public data and fields; weighted into the composite. Reference only.
Klike is a testing platform built for AI development workflows, positioned as an “AI-native testing platform.” Its core idea is not to make developers hand-write large numbers of test scripts, but to let Claude, Cursor, or any MCP Agent create, run, diagnose, and fix tests through MCP tools. The platform covers API, browser, visual regression, and load testing, and supports triggering runs from the terminal, an Agent, or a Cron Job.
Functionally, Klike provides 26 MCP tools for test creation, campaign runs, failure diagnosis, and fixing broken selectors. Browser automation relies on a Chrome Extension, supports 49 step types, and emphasizes auto-waiting to avoid brittle fixed timeouts. Visual regression supports pixel-level screenshot diffs, with changes highlighted in red. Self-healing selectors can explore the page after DOM changes, such as modified buttons, and suggest fixes. Load testing supports concurrent HTTP requests and p50/p95/p99 latency metrics, and can run from the cloud or browser side to cover localhost scenarios.
For integrations, the text explicitly mentions Claude, Cursor, any MCP Agent, GitHub, Cron, CI/CD webhooks, and a Chrome Extension. MCP onboarding includes command examples for adding the Klike MCP Server via a Bearer API Key. Pricing only discloses “Free to start, No credit card required,” without specifying free-tier limits, paid plans, or enterprise pricing. In terms of documentation, the captured content includes clear product explanations and a getting-started flow, but lacks complete API/SDK documentation, permissions, security details, usage limits, and troubleshooting information.
Klike’s strength is its deep integration between test automation and AI Agents, lowering the barrier to building automated testing from scratch. It also covers API, E2E, visual regression, and load testing, making it suitable for small teams that want to quickly complete their testing pipeline. Its downside is the lack of disclosure around key information: whether it is open source, whether self-hosting is supported, data security, SLA, payment methods, and plan limits are all unspecified. It is best suited for developers already using Claude/Cursor/MCP, early-stage product teams, and web teams that want to quickly set up regression testing.
Access from mainland China is not mentioned in the text, so the availability of klike.fr, the MCP endpoint, and the Chrome Extension needs to be tested in practice; payment methods are also unknown. If access or compliance is restricted, alternatives include Playwright, Cypress, or Selenium for browser automation, Postman for API testing, k6 for load testing, and Percy or BrowserStack for visual and cross-browser testing.
⚠ This review is compiled from public sources and does not constitute a purchase recommendation. Verify all facts on the vendor's official site. Verify on klike.fr official site.
klike.fr is an France AI Apps provider. TG4G tracks its product information, an overall rating of 8.0/10, and a China-accessibility score of China direct-connect friendly. Click "Visit Official Site" to reach klike.fr directly.