Coming soon · Plain English in, real Playwright out

Tests that write themselves.
Load. Run. Watch it live.

Describe what to check in plain English. TestVibe's AI writes the Playwright code and runs it in isolated cloud sandboxes — then load-tests the same flows and streams live telemetry from your running servers. Every run, trace, and chart in one place.

Watch the 90-second demo ↓

Real Playwright code you can read·Isolated cloud sandboxes·Traces, videos & live telemetry

Product tour

Click through the real app.

// Loading the interactive tour…
How it works

Describe it. Generation does the rest.

Three steps from a plain-English sentence to a deterministic Playwright suite you can run, load-test, and watch live.

01

Describe it

Type plain English, record yourself clicking in your browser, or point the AI at your URL and let it explore your app and suggest what to test. Everything lands as clean Gherkin .feature files.

02

AI writes the code

In an isolated cloud sandbox, the agent generates Playwright, probes your real app, and self-corrects until the script is green — powered by frontier models, no API keys to bring.

03

Run, load, observe

Deterministic Playwright — no AI tokens per run. Watch steps tick green live, open the trace, replay the video — then load test the same flows while your servers stream telemetry.

▸ Gherkin · Cart removal
The TestVibe feature editor mid-generation — the Cart removal Gherkin being written, with a 'TestVibe is writing this file…' banner over the Feature, Background and Scenario steps
▸ Code · Add item to cart.spec.js
The Code tab of a generated feature — the Add item to cart.spec.js Playwright spec the AI wrote, with an exported background and a scenarios array

The same feature in the real editor, captured mid-generation — Gherkin tab in, Code tab out.

Capabilities

Everything around the tests.

Generation is just the start — the runs, the load, the telemetry, and the workflow around them live in the same place.

Load testing

Replay your feature journeys as concurrent browser users, or drive protocol-level k6 traffic with ramp stages and pass/fail threshold gates — latency percentiles, throughput, and error rate, charted live.

TestVibe Load Testing — concurrent browser users replaying a feature with duration percentile charts

Live app telemetry

Drop a lightweight agent into any .NET or Node app and watch CPU, memory, sessions, threads, GC heap — plus any custom metric you publish — stream live while you test.

TestVibe Telemetry — a live server detail with CPU, memory, sessions and GC heap charts

AI exploration

Give it a URL. The agent browses your app in a sandbox, screenshots what it finds, and proposes the features worth testing.

TestVibe Assistant — a chat thread where the agent summarizes what the project tests

Traces & replays

Every run keeps its Playwright trace, screenshots, and video — open them right in the app, no digging through CI logs.

TestVibe Run detail — status donut, slowest tests, per-scenario results with images, videos, network and console tabs

Record in browser

Click through a flow once in your own browser — TestVibe turns the recording into a scenario, ready to generate.

Coverage

See which parts of your app have tests and which don't — the gaps are obvious.

Automations

Run your suite or a load test on a schedule, or when an event fires — results land in the same dashboards.

Test behind the wall

Staging on localhost or behind a VPN? A secure tunnel lets cloud sandboxes reach apps the internet can't.

CLI & MCP

Drive everything from the terminal or your AI tools: a testvibe CLI plus an MCP server for agents and CI.

Plugins

Extend your Gherkin with packaged steps: assert with an LLM, post to Slack, file a Jira issue, or compare visual snapshots.

Team workspaces

Organizations, roles, and invites built in. Share projects, runs, and dashboards with the whole team.

Git sync, both ways

Tests sync two-way with GitHub, GitLab, or Azure DevOps — edit in either place, changes flow both directions.

Explore the docs →

Deployment

Deploy it your way.

Recommended
Fully managed

Cloud-hosted

Sign in and start testing in seconds — we run the sandboxes, store every trace, and meter the AI. Nothing to install.

  • Isolated cloud sandboxes, warm and ready
  • First-party AI built in — no API keys to bring
  • Usage-based credits, scale up or down anytime
  • Automatic updates, backups, and security patches
Get early access →
Self-hosted

On-premise

Run the whole platform inside your own network or private cloud — for data-residency, compliance, and air-gapped teams.

  • Ships as containers — Docker Compose or Kubernetes
  • Your code and data never leave your environment
  • Bring your own AI endpoint or keys
  • Annual license · runs fully air-gapped
Talk to us about on-prem →

Same platform, same features — your choice of where it runs.

Documentation

The handbook is already written.

Quickstart Your first green run The fastest path from signing in to reviewing the result of your first generated Playwright test. docs.testvibe.com/quickstart Create tests The Test Suite Gherkin feature files, groups, the in-browser recorder, prerequisites, and importing existing tests. docs.testvibe.com/create-tests Generate Generate with AI How generation works, choosing an agent and model, reviewing progress, and reading generated output. docs.testvibe.com/ai-generation Run Run tests One test, a group, or the full suite — browsers, viewports, schedules, and live run status. docs.testvibe.com/run-tests Review Review results Read outcomes and use screenshots, videos, traces, console and network details to see what happened. docs.testvibe.com/results

Full documentation at docs.testvibe.com

Early access

Ready for tests that write themselves?

Join the waitlist — no card, unsubscribe anytime.

Get early access Read the docs

Built for SaaS & web-app teams, QA & SDET squads, platform & DevOps engineers, dev agencies, e-commerce, and fintech & regulated industries.