GLM-5.2 Reads 3-Year Medical Record Locally
MaziyarPanahi · x · 2026-07-15
The author shares how they successfully used GLM-5.2 to process a 3-year patient medical record, with the raw data strictly confined to the local Bonsai 27B model.
Key details include:
- All 292 visit notes are stored locally in Bonsai on a Mac Studio.
- The tech stack consists of llama.cpp, Metal, and ternary; the model size is 7.2GB under the Apache-2.0 license.
- The medical records never left the local machine. GLM-5.2 could only ask questions, which Bonsai answered sequentially in about 2 seconds each.
- The model eventually identified a risk factor hidden for 17 months: the combination of metformin and iodinated contrast given an eGFR of 39.
The author also mentions that a 1-bit version could reportedly shrink the model to 3.9GB, allowing it to fit onto an iPhone 17 Pro Max. The main highlight is this local privacy architecture where the orchestrator never touches the raw data.
Related event: Mac Studio Runs 27B Model Locally to Process 3-Year Medical Records(2 posts)→
More from Models
- GPT-6 Astra beats Factorio with enemies in 44 in-game hours at ~$4,500 API cost — liminal_bardo · 2026-09-11
- giffmana: the env being used in training is part of the point — giffmana · 2026-09-11
- awesome-llm-leaderboards: an open-source directory of LLM leaderboards, pricing tables, comparison tools — Last_Establishment_1 · 2026-09-11
- Anthropic claims it works to keep eval environments unidentifiable to models — MaxKannen · 2026-09-11
- Nex N2.5 Pro released on Hugging Face with 407GB of weights — jinnyjuice · 2026-09-11
- RoMa v2 image matching model unveiled in the usual black poster — ducha_aiki · 2026-09-11