AI
Make AI Content Extractor
10 min
make ai content extractor is currently in open beta it is available to all customers on paid plans, with the exception of the enterprise plan for now as a beta feature, both product functionality and pricing may change make ai content extractor is a built in app that extracts structured text and metadata from files like pdfs, word documents, images, and audio recordings—directly within your make scenarios it also converts unstructured inputs into a clean format suitable for ai apps, all without relying on third party services make ai content extractor ensures consistent output that integrates seamlessly with ai modules and ai based automations, making it ideal for tasks like pulling invoice fields from a scanned pdf, transcribing a voice note, or analyzing whatsapp image attachments key benefits no external accounts no need to connect or manage a separate account everything works within your existing make environment privacy focused your data is never stored or used for ai training instant setup start using the app immediately in your scenarios there is no need to create any connection structured output no matter the file type or layout, the app returns your data in the same, perfectly organized format every time make ai content extractor modules all modules support two ways to upload a file via a publicly accessible url mapping a file from your cloud storage if your file is stored on a cloud service like google drive, onedrive, or any other, use a module to download a file first (e g , google drive download a file) next, add a required module from make ai content extractor , and choose to extract by file alternatively, you can just paste a publicly accessible url into the module and extract information from the file that way you can use the following types of modules to build your {{scenario plural lowercase}} the current features and functionality are still in beta and are actively being improved as such, pricing is subject to change and will be updated as the app evolves document extract information from a document extracts details from a general document field description extract by select how you want to provide the content for extraction file url file select or map the file you want to extract data from file name enter (map) the name of the file you want to extract data enter (map) the file data limitation for the data for pdf and tiff, up to 2,000 pages can be processed the maximum file size is 500 mb image dimensions must be between 50 x 50 and 10,000 x 10,000 pixels if your pdfs are password locked, you must remove the lock before submission page ranges enter specific page numbers or ranges for text extraction only works if the document type is general document for example 1,3 6 limit enter the maximum number of pages to return if page ranges is defined, the limit applies only within that range leave empty to return all available pages this example shows the output (extracted information) generated from the provided pdf file extract information from an invoice extracts the values from the invoice field description extract by select how you want to provide the invoice file file url supported formats pdf , jpeg/jpg , png , bmp , tiff , heif , docx , xlsx , pptx , html file select or map the file you want to extract data from file name enter (map) the name of the file you want to extract data enter (map) the file data limitation for the data for pdf and tiff, up to 2,000 pages can be processed the maximum file size is 500 mb image dimensions must be between 50 x 50 and 10,000 x 10,000 pixels if your pdfs are password locked, you must remove the lock before submission limit enter the maximum number of values to return leave this field empty to return all recognized values on the invoice provided this example shows the output (extracted invoice details) generated from the provided invoice extract information from a receipt extracts the values from the receipt field description extract by select how you want to provide the receipt file url supported formats pdf , jpeg/jpg , png , bmp , tiff , heif , docx , xlsx , pptx , html file select or map the file you want to extract data from file name enter (map) the name of the file you want to extract data enter (map) the file data limitation for the data for pdf and tiff, up to 2,000 pages can be processed the maximum file size is 500 mb image dimensions must be between 50 x 50 and 10,000 x 10,000 pixels if your pdfs are password locked, you must remove the lock before submission limit enter the maximum number of values to return leave this empty to return all recognized values on the receipt provided this example shows the output (extracted receipt details) generated from the provided receipt image generate a c aption for an image describes the image content with a complete sentence field description caption by select how you want to provide the image file url file select or map the file you want to caption file name enter (map) the name of the file you want to caption data enter (map) the file data the file size of the image must be less than 20 megabytes (mb) the dimensions of the image must be greater than 50 x 50 pixels and less than 16,000 x 16,000 pixels image url enter a public url to the image you want to caption supported formats jpeg/jpg , png , gif , bmp , webp , ico , tiff , mpo the file size of the image must be less than 20 megabytes (mb) the dimensions of the image must be greater than 50 x 50 pixels and less than 16,000 x 16,000 pixels gender neutral caption select whether to generate captions using gender neutral terms for example, when you select gender neutral captions, terms like woman or man are replaced with person , and boy or girl are replaced with child this example shows the output (image caption) generated from the provided image url describe an image describes an image in details field description caption by select how you want to provide the image file url file select or map the file you want to describe file name enter (map) the name of the file you want to caption data enter (map) the file data the file size of the image must be less than 4 mb the dimensions of the image must be less than 33 megapixels image url enter a public url to the image you want to caption the file size of the image must be less than 20 mb the dimensions of the image must be less than 33 megapixels system prompt enter a custom instruction to guide how the ai should describe the image temperature adjust how creative or predictable the response is lower values make the output more focused, while higher values make it more diverse must be lower than or equal to 1 top p limit the response to the most likely words a lower value makes the output more focused by narrowing the range of possible words must be lower than or equal to 1 this example shows the output (image description) generated from the provided image url extract texts from a photo extracts printed and handwritten style text from a photo field description extract by select how you want to provide the photo file url file select or map the file you want to describe file name enter (map) the name of the photo you want to extract text from data enter (map) the file data the file size of the photo must be less than 20 megabytes (mb) the dimensions of the photo must be greater than 50 x 50 pixels and less than 16,000 x 16,000 pixels url enter a public url to the photo you want to extract text from supported formats jpeg/jpg , png , gif , bmp , webp , ico , tiff , mpo the file size of the photo must be less than 20 megabytes (mb) the dimensions of the photo must be greater than 50 x 50 pixels and less than 16,000 x 16,000 pixels limit enter the maximum number of text blocks to return leave this empty to return all recognized text blocks in the photo provided this example shows the output (text) generated from the provided image url get image tags retrieves a list of words related to the image field description tag by select how you want to provide the image file url file select or map the image you want to receive tags from file name enter (map) the name of the image you want to receive tags from data enter (map) the file data the file size of the photo must be less than 20 megabytes (mb) the dimensions of the photo must be greater than 50 x 50 pixels and less than 16,000 x 16,000 pixels image url enter a public url to the image you want to receive tags from supported formats jpeg/jpg , png , gif , bmp , webp , ico , tiff , mpo the file size of the image must be less than 20 megabytes (mb) the dimensions of the image must be greater than 50 x 50 pixels and less than 16,000 x 16,000 pixels limit enter the maximum number of returned tags leave this empty to return all recognized tags in the image provided this example shows the output (image tags) generated from the provided image url detect objects in an image detects various objects within an image, including the approximate location field description tag by select how you want to provide the image file url file select or map the image you want to detect objects from file name enter (map) the name of the image you want to detect objects from data enter (map) the file data the file size of the photo must be less than 20 megabytes (mb) the dimensions of the photo must be greater than 50 x 50 pixels and less than 16,000 x 16,000 pixels image url enter a public url to the image you want to detect objects from supported formats jpeg/jpg , png , gif , bmp , webp , ico , tiff , mpo the file size of the image must be less than 20 megabytes (mb) the dimensions of the image must be greater than 50 x 50 pixels and less than 16,000 x 16,000 pixels limit enter the maximum number of returned tags leave this empty to return all recognized objects on the image provided this example shows the output (objects in an image) generated from the provided image url speech transcribe an audio file transcribes an audio file field description transcribe by select how you want to provide the audio file url file select or map the audio file you want to transcribe file name enter (map) the name of the audio file you want to transcribe data enter (map) the file data an audio file (less than 2 hours long and less than 300 mb in size) supported audio formats and codecs wav , mp3 , opus/ogg , flac , wma , aac , alaw in wav container, mulaw in wav container, amr , webm , speex locales select one or more expected languages for the audio profanity filter mode choose how the transcription handles profane language channels enter (map) the indices of audio channels to transcribe separately up to two channels are supported unless diarization is enabled by default, this tool merges all input channels into a single channel and then performs the transcription if this isn't desirable, channels can be transcribed independently without merging diarization enable this to let the tool identify and distinguish between different speakers if you enable diarization, you can also set the expected maximum number of speakers this example shows the output (audio transcript) generated from the provided audio url translate an audio file translates an audio file to english field description translate by select how you want to provide the audio file url file select or map the audio file you want to translate file name enter (map) the name of the audio file you want to translate data enter (map) the file data it supports audio files less than 25 mb in size supported audio formats flac , mp3 , mp4 , mpeg , mpga , m4a , ogg , wav , webm audio url enter a public url to the audio you want to translate it supports audio files 100 mb in size supported audio formats flac , mp3 , mp4 , mpeg , mpga , m4a , ogg , wav , webm prompt enter a prompt to guide the model's style or specify how to spell unfamiliar words limited to 224 tokens temperature enter the randomness of the translation it adjusts the likelihood of the model selecting less probable words or phrases when generating text for translations, we recommend the default value of 0 must be lower than or equal to 1 this example shows the output (translation) generated from the provided audio url templates you can look for make ai content extractor templates in make's template gallery , where you'll find thousands of pre created {{scenario plural lowercase}}