Ocr4AI converts images and PDFs into Markdown or plain text so an AI app can read them. Use it in the browser (sign in with Google) or through an OpenAI-compatible API with two modes, Accurate and Fast. It supports OCR in 80+ languages, including English, Vietnamese, Chinese, Japanese, Korean and Hindi. It is pay as you go: the site lists $0.08 per million input tokens in Accurate mode and $0.02 in Fast mode, and new accounts get $0.10 of welcome credit once.