# Magnitude AI UI testing framework. Source: [Console review]() Website: [https://magnitude.run/]() Profile updated: 2026-10-16 Category: [Coding & workflows]() ## Supported tech stack Research checked: 2026-10-11 ### Runs on - Linux (Operating system) - Windows (Operating system) - macOS (Operating system) Source: [Research source]() Documented compatibility; coverage may be incomplete. Written in describes implementation, not user-project language support. Official operating-system support lists macOS, Linux and Windows. ## Console’s review — 2025-05-01 Reviewed: 2025-05-01 Author: [David Mytton]() Citation: [This review]() ### What we like Define test steps in natural language so they can adapt to interface changes. Assertions are also natural language. Test runs are cached for fast repeat execution. Data can be provided for steps e.g. login credentials. Can run anywhere you’d run Playwright e.g. in CI. Open source. ### Tradeoffs Requires two LLMs (a general model to plan and a visual model to inspect the steps). Only supports one visual model, but it can be self-hosted. This review reflects Console’s assessment at publication on 2025-05-01. Features and limitations may have changed since the review date. ## Is Magnitude a fit? Based on [Console’s review published on 2025-05-01](). Summary updated: 2026-10-16 ### Who it’s for Developers who need to test UI with adaptable test steps should consider Magnitude ### Why consider it Magnitude allows defining test steps in natural language for adaptability to interface changes ### Keep in mind Magnitude requires two LLMs, limiting its accessibility [Selection criteria]() · [About Console]() · [Browse tools]()