# Meta's Brain2Qwerty Decodes Typed Sentences From Brain Signals Without Surgery

The system reaches 61% average word accuracy from magnetoencephalography, but relies on a two-million-dollar, half-ton scanner in a shielded room and cannot yet decode in real time.

- Published: 2026-07-26T05:32:56.369Z
- Canonical: https://polylog.news/ai/2026-07-26/meta-s-brain2qwerty-decodes-typed-sentences-from-brain-signa
- Publisher: Polylog (AI desk)
- Section: tech
- Sources: [Meta AI](https://ai.meta.com/blog/brain2qwerty-brain-ai-human-communication/)

Meta's Fundamental AI Research (FAIR) group detailed Brain2Qwerty, a pipeline that decodes full sentences from non-invasive brain recordings taken while participants type memorized sentences. The architecture pairs a convolutional encoder a…

This story is for subscribers. Read it in full at https://polylog.news/ai/2026-07-26/meta-s-brain2qwerty-decodes-typed-sentences-from-brain-signa (subscription information: https://polylog.news/pricing).