The software uses proprietary voice-tracking to compare your pronunciation to a native speaker, giving you instant feedback on your accent. Milestones:

There are no explicit grammar charts or explanations in the software. Users are expected to "intuit" rules like gendered pronouns and verb conjugations through observation and repetition.

You see four images. You hear a native speaker describe one of them. You must click the correct image. The theory is that this mimics how a child learns their first language—by associating words with visual contexts rather than translating them back to your native tongue.

The "Complete" set covers the full spectrum from absolute beginner to upper-intermediate. Here is what you can expect at each stage: