# Brain2Qwerty v2 – Meta's non-invasive brain-computer interface > Source: > Published: 2026-06-29 23:42:46+00:00 The Challenge Every year, thousands of people lose the ability to speak after a stroke, an accident, or a brain disorder. While communication can be restored with [brain implants](https://www.nature.com/articles/s41586-023-06377-x) on the motor cortex, such [neuroprostheses](https://www.nature.com/articles/s41583-024-00819-9) require open-brain surgery. Today, we introduce **Brain2Qwerty v2**: a model to decode natural sentences from non-invasive magnetoencephalography (MEG) recordings. Architecture For this, we built off the Brain2Qwerty v1 architecture released [last year](https://ai.meta.com/blog/brain-ai-research-human-communication/) and recently accepted at [ Nature Neuroscience](https://www.nature.com/articles/s41593-026-02303-2). Brain2Qwerty v1 consisted of predicting keystrokes from MEG brain activity patterns recorded at Brain2Qwerty v2 overcomes this limit and generates the sentences directly from a continuous recording of brain activity. This new model now combines three hierarchical modules to jointly improve the decoding of letters, words, and sentences. Performance Trained on **10× more data** per participant, Brain2Qwerty v2 can decode complete and meaningful sentences solely from MEG signals of healthy volunteers, and reaches up to **78% word accuracy** for the best participant. Explore results Explore examples of the sentences decoded by Brain2Qwerty v2 from the brain recordings of three subjects. Remaining challenges Two major challenges remain before this method can be hoped to transfer to the clinic. First, decoding performance is not yet good enough for everyday use: this method still makes too many word-level or character-level errors to be practical. Still, our results follow a **scaling law**: the more data used for training, the better the decoder, without -- at least for now -- a detectable performance plateau. We are hopeful that larger datasets will thus further improve decoding and reduce the remaining gap with invasive neuroprostheses. Second, the MEG device used in the presented study consists of a large [scanner](https://megin.com/) -- i.e. a setup inaccessible to most patients. However, we are optimistic about this challenge: MEG sensors continue to improve, and some are now [wearable](https://www.nature.com/articles/nature26147), and could thus be more practical in the clinics. Open Science As always, we're thrilled to share our publications, code and data: