Close Menu
Versa AI hub
  • AI Ethics
  • AI Legislation
  • Business
  • Cybersecurity
  • Media and Entertainment
  • Content Creation
  • Art Generation
  • Research
  • Tools
  • Resources

Subscribe to Updates

Subscribe to our newsletter and stay updated with the latest news and exclusive offers.

What's Hot

Aprilel-1.6-15b-Thinker: Cost-effective frontier multimodal performance

December 11, 2025

Gemini 3 for developers: new inference, agent features

December 10, 2025

Anifun vs NovelAI: Which anime AI art generator is better for story creation?

December 10, 2025
Facebook X (Twitter) Instagram
Versa AI hubVersa AI hub
Thursday, December 11
Facebook X (Twitter) Instagram
Login
  • AI Ethics
  • AI Legislation
  • Business
  • Cybersecurity
  • Media and Entertainment
  • Content Creation
  • Art Generation
  • Research
  • Tools
  • Resources
Versa AI hub
Home»Tools»Unleash the power of images with AI sheets
Tools

Unleash the power of images with AI sheets

versatileaiBy versatileaiOctober 21, 2025No Comments7 Mins Read
Share Facebook Twitter Pinterest LinkedIn Tumblr Reddit Telegram Email
#image_title
Share
Facebook Twitter LinkedIn Pinterest Email

🧭TL;DR: Hugging Face AI Sheets is an open-source tool for supercharging datasets with AI models, with no code required. Vision support was added. Extract data from images (receipts, documents), generate visuals from text, and edit images, all in spreadsheets. Leverage thousands of open models through inference providers.

Analyze images with AI sheets

We are excited to release a major update to Hugging Face AI Sheets, an open source tool for building, transforming, and enriching data using open AI models. AI Sheet is powered by inference providers, so you can use thousands of open models powered by the best inference providers on the planet.

The first version of AI Sheets made it easy to structure and enrich your text content. We are currently adding vision to the AI ​​sheet.

Images are everywhere: product photos, receipts, screenshots, diagrams, graphs, logos. These documents contain structured information waiting to be extracted, analyzed, and transformed. Today, you can finally interact with visual content directly in your AI sheets. View images, analyze them, extract information, generate new images, and even edit them in real time, all within the same workflow.

Your images have a story to tell

Images contain valuable information such as product catalogs, support tickets, research archives, receipts, documents, and more. You can now upload images directly, use datasets containing images, and use vision models to extract, analyze, and structure information within images.

What you can do:

Image description and classification – Generate captions for product photos, classify document types, or tag images by content Structured data extraction – Extract items from receipts, data from charts, or text from scanned documents Add context and metadata – Automatically label images with relevant attributes, quality scores, or custom annotations

Just like with text columns, you can iterate through the prompts, manually edit the output, and give the thumbs up to tell the model what you want. Your feedback will give us some examples to create better results.

Examples: From receipts to structured expenses

Imagine you come back from a trip with a stack of receipts. Upload these to your AI sheet and create a column with prompts like this: Extract the seller name, date, total amount, and expense category from this receipt.

The AI ​​sheet processes each receipt and provides a clean table with all details extracted. You can edit mistakes, evaluate and verify good results, and regenerate the rest to improve. Export the final dataset as CSV or Parquet for expense tracking tools.

Or maybe you’re digitizing handwritten recipes from old family notebooks. Create columns to extract ingredients, cooking times, and dish types to turn your personal archives into searchable, structured data sets.

Generate and transform text and images in the same flow

Need visuals for your content? AI Sheets uses AI models to generate and edit images directly within your spreadsheet, keeping your entire content creation workflow in one place.
What you can do:

Generate images from text – Create social media graphics, thumbnails, or illustrations that match your content Edit and transform existing images – Modify uploaded images or generated visuals – Change styles, add elements, adjust composition Create large variations – Generate multiple versions or styles to test what resonates with your audience Build a visual content library – Create consistent brand assets across large content campaigns

Example: Creating a content calendar using visuals
Imagine you plan to post about healthy recipes on social media for a month. I have a spreadsheet with post titles and descriptions, but no images yet.

Create an image column with a prompt like this: Generate an appetizing food photo for {{title}}. Style: Bright, overhead shots, natural light.

AI Sheet generates a unique image for each post. Is it incorrect? Create another column and edit it. Convert the image to have a rustic wooden background and add fresh herbs as garnish.

You can iterate on generating and editing prompts to try different approaches. Your entire content calendar (copy and visuals) is saved in one spreadsheet and can be scheduled or exported instantly.

step-by-step guide

Now let’s take a look at the AI ​​sheet in action. We use an open model to unlock the knowledge within handwritten recipes, like the ones you might find from your grandma.

Upload data

There is a folder with photos that you can easily upload to the app.

upload

The result will be a spreadsheet like this:

table

Understand how AI works

Each column in your spreadsheet can be transformed, extracted, queried, and anything else you can imagine using AI actions.

To see this in action, click the overlay at the top of the column.

AI action

Image columns come with image operations such as text extraction, image interrogation, object detection, colorization, adding text, and possible custom actions.

Text columns include summaries, keyword extraction, translations, and custom actions.

Prompts and models define every AI action. Let’s see what you can do with a handwritten recipe dataset.

Extract text from images.

AI sheets come with templates for extracting text from images.

text extraction

The result of this action is an AI-generated column containing the transcribed text. Let’s take an example:

recipe

For the image above, the extracted text would look like this:

Notes: From 1 box Duncan Hines Yellow Cake Mix 1 box instant lemon pudding 2/3 cup water 1/2 cup mozola oil 4 eggs Lemon flavor, optional. Place in a mixing bowl and mix for 10 minutes. Remember…For quality printing, call or write Gatling & Pierce PRINTERS at 332-2579. We have 22 years of service in Northeastern Carolina.

Not bad! But I’m interested in the recipe text because I see that it includes print text for the header and footer. This text is included because we are using the default template for text extraction, which looks like this:

Extract and transcribe all visible text from images, including signs, labels, documents, or written content.

Try out custom prompts.

custom

The extracted recipe details are as follows:

– 1 box of Duncan Hines Yellow Cake Mix
– 1 box of instant lemon pudding
– 2/3 cup water
– 1/2 cup mazola oil
– 4 eggs
– Lemon flavor according to taste
– Place in a mixing bowl and mix for 10 minutes

This is amazing! But what about more complex images? By default, AI Sheets uses models that balance speed and accuracy, but you can experiment with thousands of models. The example above uses the default vision language model Qwen/Qwen2.5-VL-7B-Instruct.

Let’s test the SoTA inference model, Qwen/Qwen3-VL-235B-A22B-Reasoning, with a more challenging image.

Quen 3

Here is a comparison between the models:

Qwen/Qwen2.5-VL-7B-Instruct Qwen/Qwen3-VL-235B-A22B-In a large bowl, combine the meat, onion, bread crumbs, 1/2 nutmeg, and cheese, and sprinkle around. Then blend – sprinkle at the end and blend again. Bake in a large pan for 10-15 minutes. Let stand for 5 minutes before eating. Mix meat, onion, bread crumbs, 1/4 nutmeg, and cheese in a lg bowl and sprinkle around. Then blend – blend the last bit of spinach again. Bake in a frying pan for 50-60 minutes. @350 – let stand for 5 minutes before eating

Both models produce very similar output, but include two subtle but important details (in bold): temperature and the main ingredient, spinach.

Clean, transform, and enhance your text

Once you are satisfied with the extracted text, you can further transform and enhance it. You need to perform an AI action using the new column, as follows:

format

You now have beautifully structured HTML pages for each recipe.

html

Edit and convert images.

Finally, AI Sheets integrates image-to-image models like Qwen-Image-Edit. This means you can perform AI actions to transform and enhance your images.

For example, let’s say you want to offer recipes and old style. You need to go to the column and use the black and white template like this:

Conversion bandwidth

result:

monochrome

Export the dataset

Once you’re happy with your new dataset, export it to your hub. If you don’t want to share it with the community, you can export it to your organization, personal profile, or make it private.

export

Check out the dataset you just created.

What’s next?

You can try out AI Sheets without installing them or downloading and deploying them locally from a GitHub repository. To run locally and get the most out of it, we recommend subscribing to PRO and getting 20x your monthly inference usage.

If you have any questions or suggestions, let us know in the Community tab or by opening an issue on GitHub.

author avatar
versatileai
See Full Bio
Share. Facebook Twitter Pinterest LinkedIn Tumblr Email
Previous ArticleVeo 3.1 model update: Enhanced realism and richer audio for creators now available via Gemini API and Google Cloud | AI News Details
Next Article California’s AI law sets national trends
versatileai

Related Posts

Tools

Aprilel-1.6-15b-Thinker: Cost-effective frontier multimodal performance

December 11, 2025
Tools

Gemini 3 for developers: new inference, agent features

December 10, 2025
Tools

Accenture and Anthropic partner to power enterprise AI integration

December 10, 2025
Add A Comment

Comments are closed.

Top Posts

New image verification feature added to Gemini app

December 7, 20256 Views

Aluminum OS is the AI-powered successor to ChromeOS

December 7, 20255 Views

UK and Germany plan to commercialize quantum supercomputing

December 5, 20255 Views
Stay In Touch
  • YouTube
  • TikTok
  • Twitter
  • Instagram
  • Threads
Latest Reviews

Subscribe to Updates

Subscribe to our newsletter and stay updated with the latest news and exclusive offers.

Most Popular

New image verification feature added to Gemini app

December 7, 20256 Views

Aluminum OS is the AI-powered successor to ChromeOS

December 7, 20255 Views

UK and Germany plan to commercialize quantum supercomputing

December 5, 20255 Views
Don't Miss

Aprilel-1.6-15b-Thinker: Cost-effective frontier multimodal performance

December 11, 2025

Gemini 3 for developers: new inference, agent features

December 10, 2025

Anifun vs NovelAI: Which anime AI art generator is better for story creation?

December 10, 2025
Service Area
X (Twitter) Instagram YouTube TikTok Threads RSS
  • About Us
  • Contact Us
  • Privacy Policy
  • Terms and Conditions
  • Disclaimer
© 2025 Versa AI Hub. All Rights Reserved.

Type above and press Enter to search. Press Esc to cancel.

Sign In or Register

Welcome Back!

Login to your account below.

Lost password?