Are you the maintainer?
Claim this page →A multi-page document understanding model whose high-resolution DocCompressor module enables efficient question answering without OCR.
01 / About
mPLUG-DocOwl2 answers questions over multi-page documents without depending on OCR, using a high-resolution DocCompressor module for efficient multi-page understanding.
Choose mPLUG-DocOwl2 for long-form PDFs such as contracts or research papers, where it delivers efficient multi-page understanding without relying on OCR.
02 / Credibility
Public-signal score, discounted for bought-star and bot-stargazer patterns — not a user rating. How scores are calculated.
Score unavailable
We haven't been able to compute a GitHub credibility score for this tool yet. It'll appear here once the evaluator runs.
03 / Discussion CREDIBILITY-GATED
Reading is open to everyone. Only verified humans or builders at GitHub B+ can post or rate — every comment carries its author's credibility.
04 / Related
ToolControl browsers programmatically with LLM agents through a high-level, LLM-friendly API.
ToolJuliusBrussee
An experimental LoRA fine-tune of Gemma to speak caveman-mode natively.
ToolJuliusBrussee
A TypeScript implementation of the caveman compression engine.
Toolegorfedorov
Tracks token usage, blocks redundant reads, and supports .contextignore and budget alerts.
ToolNodeNestor
A proxy plugin that rolls context compression past 100K tokens.
Toolnadimtuhin
Restructures CLAUDE.md and docs for roughly 90% token savings with a CLI audit and compress.
05 / Build
Browse the catalogue for harnesses, tools, and blueprints — each scored on real GitHub credibility.
Browse the catalogue