MonkeyTest

Visit Website

Introduction

MonkeyTest is an AI exploratory testing agent for websites. Its homepage says its agents use your site like a real user to find bugs, report them, and verify them once fixed, with the tagline “Break your site before your users do”. It targets teams shipping fast, including with AI-generated code, whose test suites miss flows such as forms that quietly return errors.

MonkeyTest is open source: the code is at monkeytestai/monkeytester on GitHub under the AGPL-3.0 licence, with docs at docs.monkeytest.ai. The latest release at review was v0.7.1 (2 November 2026), and the repository had commits in November 2026, although it is a young project with few stars. You can run it yourself with your own LLM provider or use a free cloud tier.

The workflow is “zero to bug report in three steps”: find (agents explore your site), prove (they capture evidence) and verify (re-check once fixed). The site stresses that it does exploratory QA rather than providing browser infrastructure, and suggests running it as a ship-check on preview deployments.

In the quality loop MonkeyTest fits the end-to-end stage as an exploratory complement to scripted tests, and can act as a release gate on previews.

Features

1. Exploratory AI agents: Agents explore sites like real users.
2. Bug reports with evidence: Reproducible reports for issues found.
3. Fix verification: Re-check bugs once fixed.
4. Self-hosted or cloud: Run it yourself or use the free cloud.
5. Bring your own LLM: Configure LLM providers.
6. Preview checks: Run against preview deployments before release.

Pros

1. Open source: AGPL-3.0, self-hostable.
2. Free options: Free self-hosting and a free cloud tier.
3. Finds unscripted bugs: Covers flows your suite never tested.

Cons

1. Young project: Early version numbers and a small community.
2. AGPL obligations: Network-use terms matter if you modify and host it.
3. LLM costs: Self-hosting uses your own LLM tokens.

Usage and Links (Example)

Follow the docs to run MonkeyTest locally or in CI with an LLM API key, point it at a preview URL, and review the bug reports it produces. Alternatively, try the free cloud on a staging site first.

Useful links:

• Documentation
• GitHub repository

Pricing

Open source under AGPL-3.0 with a free cloud option; see site for any paid tiers.

Recommended for?

• Teams shipping fast that want exploratory coverage
• Developers checking preview deployments
• Teams experimenting with AI QA agents

Conclusion

MonkeyTest is a promising open-source AI exploratory testing agent. It is young, so pair it with scripted tests and review its findings, but it is easy to try for free.