{"name":"mafdet","repo":"Mafdet-AI/mafdet","url":"https://midorreal.com/project/thelastoxygen-mafdet","repository":"https://github.com/Mafdet-AI/mafdet","description":"You type. It already knows. Local LLMs that read and think while you type.","facts":{"license":{"value":"Apache-2.0","source":"https://github.com/Mafdet-AI/mafdet","checked":"22 Nov 2026"},"language":{"value":"Python","source":"https://github.com/Mafdet-AI/mafdet","checked":"22 Nov 2026"},"last_release":{"value":"v0.6.0, 14 Nov 2026","source":"https://github.com/Mafdet-AI/mafdet/releases","checked":"22 Nov 2026"},"releases":{"value":"7","source":"https://github.com/Mafdet-AI/mafdet/releases","checked":"21 Nov 2026"},"contributors":{"value":"5","source":"https://github.com/Mafdet-AI/mafdet/graphs/contributors","checked":"21 Nov 2026"},"archived":{"value":"no","source":"https://github.com/Mafdet-AI/mafdet","checked":"22 Nov 2026"}},"stars":81.0,"writeup":{"what_it_is":"MAFDET is a local app that runs LLMs and processes the user's message while they type it. It reads the text ahead of time and can start reasoning about it before the user presses Enter. It has a window and a terminal interface, and runs on llama.cpp.","audience":"People running local LLMs on consumer GPUs or CPUs who want shorter waits after sending a message","claims":[{"kind":"specific","claim":"Reduces the mean wait after Enter on Qwen3-4B from 12.2 s to 3.2 s on 24 word problems","excerpt":"the mean wait after Enter goes from **12.2 s to 3.2 s**","status":"not_checked"},{"kind":"specific","claim":"Reduces the mean wait after Enter on Qwen3-14B from 108 s to 47 s","excerpt":"on Qwen3-14B, from **108 s to 47 s**","status":"not_checked"},{"kind":"specific","claim":"Reference machine is an RTX 2060 with 6 GB of VRAM, and it also runs on CPU only","excerpt":"The reference machine is an **RTX 2060 with 6 GB of VRAM**, and everything also runs on **CPU only**.","status":"not_checked"},{"kind":"specific","claim":"Runs on Windows with an NVIDIA GPU or a CPU, with Linux not yet tested","excerpt":"**Linux** has its llama.cpp builds and code paths in place, but has not been tested yet","status":"not_checked"},{"kind":"specific","claim":"Speed estimates are within 12% of measured speeds on the reference PC","excerpt":"is within 12% of the measured speeds on the reference PC","status":"not_checked"}],"alternatives":[],"written_by":"AI, from the project's README","date":"2026-10-25"},"why_now":null,"owner_supplied":null,"score":{"verdict":"sleeper","hype":19,"reality":85,"gap":-66,"momentum":null,"confidence":60,"version":"score-2.0","rule":"verdict-2.0","calculated_at":"2026-11-22T07:00:00+00:00","verdict_source":"rule"},"early_read":null,"changes":[{"date":"2026-11-14T11:43:05+00:00","text":"Latest release v0.6.0","source":"https://github.com/Mafdet-AI/mafdet/releases"},{"date":"2026-10-22T14:38:05+00:00","text":"First release","source":"https://github.com/Mafdet-AI/mafdet/releases"},{"date":"2026-10-21T08:37:56+00:00","text":"Repository created","source":"https://github.com/Mafdet-AI/mafdet"}],"cite":"mafdet is a GitHub project at https://github.com/Mafdet-AI/mafdet written mainly in Python, described there as \"You type. It already knows. Local LLMs that read and think while you type\".","attribution":"Data from Mid or Real, https://midorreal.com/project/thelastoxygen-mafdet, checked 22 November 2026. Attribution with a link is required.","still_real":null}