snipmd, an offline Mathpix Snip alternative for macOS

Описание и применение

snipmd — бесплатный офлайн‑заместитель Mathpix Snip для macOS, использующий модель GLM‑OCR для быстрого преобразования изображений уравнений в LaTeX, Markdown, CSV или обычный текст.

Задача

Need to extract LaTeX from equations in PDFs or slides quickly without paid services or internet dependency.

Результат

snipmd matched 48 benchmark equations with 0.5 s median latency, outperforming Mathpix Snip, LaTeX‑OCR, and Pix2Text on the same test set.

Рабочий процесс

• 1. Press global hotkey to activate macOS crosshair selector.
• 2. Drag to crop the equation image; crop is flattened onto white and padded.
• 3. GLM‑OCR (0.9B model) processes the image locally via MLX on Apple Silicon or Ollama on other machines.
• 4. Post‑processing strips code fences, normalizes delimiters, removes odd token spacing, and converts tables to Markdown/CSV.
• 5. Result is copied to clipboard, shown in a notification, and saved to a local history file.

Доказательства

I kept running into the same small task: an equation sits in a PDF or a lecture slide, and I need it as LaTeX to paste into my notes or a paper.

Источник

Mathpix Snip solves this well. Press a hotkey, drag a box, get LaTeX. The problem is the free plan stops at 10 snips a month, Pro costs 4.99 dollars a month, and every snip goes to their servers.

Источник

snipmd matched all 48 after normalizing for equivalent notation, such as pmatrix versus a raw array. Median latency was 0.5 seconds from mouse release to clipboard, with the model already loaded.

Источник

snipmd is free, MIT licensed, and runs entirely on your machine. The code, the benchmark, and a comparison with Mathpix Snip, LaTeX-OCR, and Pix2Text are on GitHub: https://github.com/Arthur031221/snipmd

Источник