Sources: creator videos at the linked timestamps. EU status is information, not legal advice; see how we check.
Brain2Qwerty v2 is a research system from Meta that decodes text a person types from non-invasive brain recordings (magnetoencephalography, MEG). It is a follow-up to the original Brain2Qwerty work, which predicted typed characters from brain activity recorded while participants typed sentences.
According to a creator summary, version 2 was trained on about ten times more data and reaches roughly 69% character accuracy and 78% word accuracy. These figures are secondhand and have not been checked against an official publication. The code is reported to be released.
This is a research project rather than a consumer product. MEG requires specialised, expensive scanning equipment in shielded rooms, so it is not usable by the general public. No official pages were available to confirm details such as the license, the dataset or the exact release date.
Watch the moment: the exact second top creators show it. Videos load only when you press play, via youtube-nocookie.com.
Decodes non-invasive MEG brain signals into typed text. Version 2 trained on 10x more data reaches about 69% character and 78% word accuracy; code is released.
Closest neighbours by meaning, not just by category.
Model · Meta
Closed-source Meta model for coding and agentic workflows with 1M-token context and multimodal input, offered via API.
App · Perplexity
Knowledge-wiki "second brain" feature in Perplexity Computer with an Obsidian-style knowledge graph.
Model
One- and two-step image generation model distilled from Z-Image-Turbo using text-stream adversarial distillation.
Hardware · GenSpark
Credit-card-sized MagSafe AI recording card from GenSpark that turns meeting recordings into notes in its Second Brain.
Model · Hugging Face
Open 14.8B language model from Hugging Face that reads raw DNA in chat and answers in English, built on Qwen3-14B.
Model · Google
Google's 2B-parameter Gemma 2 instruction-tuned model, repackaged as a text encoder for PixelDiT and PiD workflows in ComfyUI.