“OCR”

Extract text from an image using the on-device OCR model via the zettos-ai engine (private/offline).

Change or evaluate Xberg OCR backends, preprocessing, caching, page acceptance, geometry, hOCR structure, table reconstruction, or cross-backend quality.

ローカル Tesseract OCR でスクリーンショット画像群を Markdown に文字起こしする。Playwright による自動スクショ取得も可能。社外秘ドキュメントを外部 API に出さずに処理したいときに使う。

Image URL → plain text OCR for bots.

Arabic-first OCR, translation and document extraction.

Read PDFs and images as markdown or text, with exact costs and hard spend caps.

macOS computer-use agent that reads the screen with OCR, uses a small classifier to choose each click and a writer only for text.

Single-binary, DeepSeek-native terminal coding agent with a code graph, local OCR, model routing and an OS sandbox.

Chinese-language tender agent platform that interprets bid documents, drafts bids, runs 21 compliance checks and formats docx output.

Moondream 3 Preview is a vision-language model from Moondream with 9B total parameters and 2B active (mixture-of-experts).

Moondream 3.1 is a vision-language model from Moondream, the successor to Moondream 3 Preview, with 9B total parameters and 2B active (mixture-of-experts).

Open-source document conversion library that turns PDFs, Office files and audio into structured data.

Context engine that indexes, retrieves and extracts entities from documents for agents.

OCR for ALL: AI Image to Text, PDF, LaTeX Math, Table & Invoice Parser

ocr-for-all

Ultra-high accuracy AI OCR.