1 / 3
【Overview】 BrowserBash is a command-line tool that turns plain-English objectives into real Chrome browser automation. You describe what you want to happen, and an AI agent figures out the clicks and typing, so there are no CSS selectors, XPath expressions, or code to maintain. It is built for AI coding agents that need to validate their own work in a real browser, as well as QA and test engineers who want readable end-to-end tests. The tool returns a structured, machine-readable verdict, making it a validation layer for AI coding agents and CI pipelines. Because assertions compile to real Playwright checks, you get deterministic evidence rather than guesswork. BrowserBash runs on free local models via Ollama or free OpenRouter models, with no API keys or credit card required. BrowserBash is aimed at teams that already use cloud browser grids like Browserbase, LambdaTest, and BrowserStack, and at platform teams that need budgets and cost controls in CI. It also works with any CDP endpoint and a real local Chrome, using the same test across providers. 【Key features】 • Plain-English objectives: Describe what you want in plain English and the AI agent handles the clicks and typing, with no CSS selectors, XPath, or code to maintain. • Markdown test files: Save scenarios as readable *_test.md files that can be composed with @import so shared setup and reusable flows stay in one place. • Variables & secret masking: Template values into tests with variables and pass credentials as secrets that are automatically masked in logs and output. • Free models, zero keys: Run on free local models with Ollama or free models on OpenRouter, with no API keys or credit card required. • Real browsers & providers: Drive a real local Chrome, any CDP endpoint, or cloud grids like Browserbase, LambdaTest, and BrowserStack with the same test. • MCP server: browserbash mcp plugs into Claude Code, Cursor, Codex, or any MCP host and exposes three tools: run an objective, run a test file, run a suite. • Deterministic assertions: Verify steps compile to real Playwright checks on URL, title, visible text, element counts, and stored values, with expected-vs-actual evidence on failure. • API steps (testmd v2): Seed data over plain HTTP and then verify through the UI, with per-step execution on one browser session. 【Use cases】 1. Write end-to-end browser tests in plain English. 2. Let a coding agent validate a change in a real browser over MCP. 3. Gate CI pipelines on browser test pass/fail exit codes. 【Pricing】 Free · $0: The full open-source CLI — every command, no locked features, plain-English automation, all browser providers (local Chrome, any CDP endpoint, Browserbase, LambdaTest, BrowserStack), and free local models via Ollama or free OpenRouter models — zero API keys. Supporter: Everything in Free, extended cloud run retention beyond 15 days (coming soon), and directly supports ongoing development of the project. 【Pros and cons】 Pros: - The plain-English approach removes the selector maintenance that makes browser tests brittle, so QA engineers can write tests without deep Playwright knowledge. - Because it runs on free local models or free OpenRouter models, teams can adopt it without API keys or a credit card. - The MCP server makes it a natural fit for AI coding agents that need to check their own changes in a real browser. Cons: - When using cloud grids, free local Ollama and OpenRouter models stay local-only. - Cloud grids use the builtin engine, which speaks the Anthropic API. - Monitoring alerts fire only on pass/fail state changes. - When a budget is exhausted, remaining tests are reported as skipped and the exit code is 2. - Playwright import is heuristic; untranslated specs go to an import report. 【FAQ】 Q: Is BrowserBash free? A: Yes, the Free plan costs $0 and includes the full open-source CLI with every command and all browser providers. A Supporter plan is also available and includes everything in Free plus extended cloud run retention beyond 15 days (coming soon). Q: Who is BrowserBash for? A: It is designed for AI coding agents that need to validate their own work in a real browser, QA and test engineers building end-to-end browser tests, and platform teams that need budgets and cost controls in CI. It also suits teams already using BrowserStack or LambdaTest grids. Q: Do I need API keys to use BrowserBash? A: No. BrowserBash runs on free local models with Ollama or free models on OpenRouter, with no API keys or credit card required.
Free · $0 (The full open-source CLI — every command, no locked features, Plain-English automation: an objective in, real browser actions out, All browser providers: local Chrome, any CDP endpoint, Browserbase, LambdaTest, BrowserStack) Supporter (Everything in Free, Extended cloud run retention beyond 15 days (coming soon), Directly supports ongoing development of the project)
WPS AI is an AI assistant built into WPS Office that helps users write documents, generate presentation outlines and…
Baidu Comate is an AI coding assistant built by Baidu on its ERNIE model. It goes beyond simple autocomplete: it…
UniConverter is an AI media conversion and enhancement tool for video, audio, and images. It combines everyday…
6 similar tools to consider instead of BrowserBash