Plot2txt creates and manages API methods for performing analysis of images and other binary documents like audio, using a combination of computer vision, machine learning and traditional OCR. These algorithms run using Amazon Web Services (AWS) Lambda functions, overall providing a powerful means for creating highly scalable computer vision applications. Input is base64 encoded PNG or WAV files, output is JSON, base64 encoded JSON, or base64 encoded gzip’d collections of text and images. Input is restricted to less than 2Mb per invocation, ie., large files are best divided and processed in parallel. Plot2txt methods are ephemeral; there is no associated storage and all temporary data is erased after each invocation, which is restricted to 30 seconds or less in duration. The language models used are English at this stage, support for other languages potentially available upon request. Please email us at email@example.com with any questions.