Fume is an AI QA team that generates and maintains Playwright end-to-end browser tests from video recordings, self-healing them as your application evolves.

How much does Fume cost?

Fume offers a Team plan at $500/month with a 30-day free trial, and an Enterprise plan with custom pricing.

Fume can record user flows from videos, generate Playwright tests, run them in the cloud or your CI, self-heal broken tests, and migrate existing Playwright, Selenium, or Cypress tests.

Fume vs AI No-Code Test Builders?

Fume is a fully autonomous QA department that designs and maintains resilient tests, handles dynamic data, and provides 100% code ownership, unlike no-code builders that require human design and often lead to vendor lock-in.

Do I own the Playwright code even after I cancel my subscription?

Yes, you always own the output Playwright code, even if you cancel your subscription.

Launching soon — get early access:

/

AI Agents

/

Fume

Autonomous testing

AI-powered self-healing

Playwright code ownership

Best for QA automation

Developer-centric

Effortless AI Testing for Web Applications

Autonomous

Testing

PAID

ADVANCED Complexity

HYBRID

Last updated: May 13, 2026

fumedev.com

Quick Answer

Fume is an autonomous AI agent that serves as a full QA team, generating and maintaining Playwright end-to-end browser tests from simple video recordings. It features AI agent fallbacks for self-healing flaky tests and integrates into CI/CD pipelines. Best for developers and QA engineers building production AI workflows. Offers a free trial for its Team Plan.

Decision Intelligence

Best For

Automating end-to-end browser testing for web applications

Replacing manual QA efforts for growing teams

Generating resilient Playwright test suites

Maintaining tests against evolving UI/UX

Not Ideal For

Non-technical users seeking a purely no-code testing solution

Teams that prefer manual test design and maintenance

Testing non-browser-based applications

Fit Score

Complexity

ADVANCED

Company Size

SMALL

MEDIUM

ENTERPRISE

Team Size

SMALL_TEAM

LARGE_TEAM

Skill Level

EXPERT

About Fume

Fume is an autonomous AI agent that acts as a full QA team, generating and maintaining Playwright end-to-end browser tests from simple video recordings. It self-heals flaky tests and integrates seamlessly into CI/CD pipelines. Own 100% of the generated Playwright code.

AI-extracted from public sources. May contain errors. Methodology · Are you Fume? Claim listing

What Makes Fume Unique

Fume uniquely combines video-based test generation with an AI agent fallback system that self-heals flaky Playwright tests, ensuring high resilience against UI/UX changes and dynamic data, while providing full code ownership.

Capabilities

Extract test cases from video recordings

Generate Playwright end-to-end tests

Self-heal flaky tests with AI agent fallbacks

Run tests in cloud or CI/CD environments

Migrate existing Playwright, Selenium, and Cypress tests

Explore applications to write new tests

Provide insights on test failures

Automate test maintenance

Task Categories

Workflow Orchestration

Pros

+Generates robust and optimized Playwright tests autonomously
+Self-heals flaky tests by agentically adapting to UI/UX changes
+Provides full ownership of generated Playwright code, preventing vendor lock-in
+Supports migration of existing Playwright, Selenium, and Cypress tests
+Offers free cloud test runners and seamless CI/CD integration
+Reduces manual test design and maintenance efforts significantly

Limitations