About

NoteLMs helps students organize study notes, and every use feeds a research comparison of language models.

Product

Built for students who want notes sorted, readable, and easy to revisit out loud.

  1. Categorizing notesPaste a note and NoteLMs assigns it to a subject so your library stays organized.
  2. Formatting notesRaw pasted text is turned into clean, structured HTML that’s easier to skim and study from.
  3. Reading aloudOn Voice, paste text and hear it read back, useful for review without staring at a screen.

Technical details

Under the hood, NoteLMs is also a live experiment in note classification.

  1. Research goalCompare zero-shot BERT, fine-tuned BERT, and a full reasoning LLM (GPT-OSS 20B) on academic subject classification.
  2. Eval dataWe’ve already run 2,000 offline evals. Every time someone uses the product, that classification is added to the research set.
  3. How categorization worksThree models vote in parallel: zero-shot BERT, fine-tuned BERT, and GPT-OSS 20B. An orchestrator (also GPT-OSS 20B) reviews those votes and chooses the final subject.
  4. How image notes workUploaded or dragged note photos are sent to OpenAI vision, which extracts the raw text; that text then goes through the same classify → format → save pipeline as pasted notes. This is a temporary feature that will be removed soon.
  5. How reading aloud worksText is split into smaller chunks, then Orpheus 3B (orpheus-3b-0.1-ft) synthesizes speech chunk by chunk so longer passages stay clear and stable.