10 Things You Can Do with the Filestack Free Plan

The Filestack free plan runs image resizing, format conversion, PDF generation, face blurring, caching, screenshots, minification, zipping and storage. Each example takes one URL. Most AI-powered operations that read and interpret a file run on the higher plans.

Every example below is live. Click any URL in this article and you will get back exactly the file shown, at exactly the size quoted.

Key takeaways

• A free key can transform, convert, deliver, archive, and store files through URL-based tasks.
• Filestack tasks chain from left to right, so remove pixels before changing appearance or encoding.
• Resizing before format conversion usually saves more bandwidth than changing the format alone.
• A store URL creates a new asset on every request, so call it once and keep the returned handle.
• Facial detection runs on the free plan, while most other AI-powered operations are available on the higher plans.

What a free key gets you

Sign up for a free account, open the developer portal, and copy the API key from the top of the dashboard. That key is all you need. There is no SDK to install and nothing to configure before the first request.

Every operation below follows the same URL shape:

https://cdn.filestackcontent.com/TASK/FILE_HANDLE

TASK is the operation and its parameters, and tasks chain left to right when you need more than one. FILE_HANDLE is the ID returned when you upload a file. Change the task, get a different file back, which is why these examples are one line each. Where a chain appears below, the order is always the same: remove pixels first, change appearance second, set the encoding last.

The handle identifies the application, so a stored file needs no API key in the URL. The one exception is a task reading from an external web address rather than a handle, which needs the key to know whose application is asking. The examples below run against files on a demo account, so the links work as you read; swap in your own handles and the same URLs run against your files.

Here is the photograph most of the examples start from. The original upload is a 1254 by 1254 PNG of 2,267,254 bytes, and it is shown here at 700 pixels wide because serving the full file to every reader would be exactly the mistake this article is about.

1. Resize an image to any width

Resizing is the operation most people reach for first, and it accepts width, height, fit and align on a free key.

https://cdn.filestackcontent.com/resize=width:300/VK2LSKTSTmuS6U5fqFkp

 

That thumbnail is 139,331 bytes against the original 2,267,254, a sixteenth of the weight, computed on the way out with no stored variant anywhere. If you want the full picture of cropping, watermarking and chaining several edits into one request, our guide to the image editing api covers it properly.

2. Convert an image to WebP

Format conversion is the same idea with a different task. Ask for WebP and you get WebP, whatever the original was. Tasks also chain, so this one resizes first and converts second.

https://cdn.filestackcontent.com/output=format:webp/resize=width:700/VK2LSKTSTmuS6U5fqFkp

 

Those are the same pixels as the photograph at the top of this article. As a PNG it weighs 721,764 bytes. As WebP it is 113,302, a saving of six sevenths for no visible difference. If you are deciding which format to standardise on, our reference on when to convert to webp and when to leave a file alone goes format by format.

3. Turn a document into a PDF

The pdfconvert task takes a document and returns a PDF with the page setup you ask for. It accepts pageformat, pageorientation and pages, so you can pull three pages out of a long file as easily as converting the whole thing.

https://cdn.filestackcontent.com/pdfconvert=pageformat:a4/jLoemkNMSiyxhKHxo4G0

That returns a 1,314 byte A4 PDF, and you can open the converted PDF to check it. Since a PDF will not display inline, here is the same document rendered to an image with output=format:png, which is another free-plan task worth knowing about:

 

This is the one to try if you generate invoices, reports or receipts and currently run a headless browser to do it.

4. Blur faces in an uploaded photo

Face detection and its blurring, pixelating and cropping variants are on the free plan. Pass an amount and the faces come back obscured, with the rest of the frame untouched.

https://cdn.filestackcontent.com/blur_faces=amount:10/6KHnf6yQmuNZPbxpaH2A

Here is a second photograph before the task runs:

 

And the same file with the task applied:

 

Nothing was marked up by hand. The face was found and obscured, while the mountains, river, and jacket came through sharp. The task also takes faces, type, and buffer for finer control. Our blur faces walkthrough has the Node.js and React code.

5. Control how long a file stays cached

Every file you deliver goes through the CDN by default, and you can set the cache window per URL.

https://cdn.filestackcontent.com/cache=expiry:3600/VK2LSKTSTmuS6U5fqFkp

There is deliberately no picture for this one. The response is the source photograph at its original 2,267,254 bytes, byte for byte identical, now cached for an hour. Caching changes delivery, not pixels, and that is the point of the task. Cache behaviour is where a free plan either saves your quota or burns it, so it is worth understanding how edge caching and responsive delivery fit together before you go live. Our guide to Filestack CDN file delivery explains the 30 day default and when to bypass it.

6. Capture a screenshot of any web page

You do not need to upload anything for this one. Give the task a URL and it renders the page and hands back an image.

https://cdn.filestackcontent.com/YOUR_API_KEY/urlscreenshot/https://example.com

Replace YOUR_API_KEY with the key for the application making the request. The task renders example.com on demand, which makes it useful for link previews, competitor snapshots or automated visual records without running a browser farm yourself. Because the output is just another file, you can chain a resize onto it and serve a thumbnail of a live page from the same request.

7. Minify a JavaScript file

Filestack handles more than images, and minification is the clearest proof of it. Point the task at a stored JavaScript file and it strips it down.

https://cdn.filestackcontent.com/minify_js/waEy1tpnTNKobiccMIA1

The stored file looks like this, at 122 bytes:

function greet(name) {
  var message = 'hello ' + name;
  if (true) {
    console.log(message);
  }
  return message;
}
greet('world');

And this is what comes back, at 74 bytes, which you can read in the browser yourself:

function greet(a){var b="hello "+a;return console.log(b),b}greet("world");

Names are shortened, the dead branch is gone, and the two statements are folded into one return. It happens at delivery time rather than in your build step, which matters most for assets you do not control, such as scripts uploaded by your own users or a partner.

8. Zip several files into one download

The zip task takes an array of handles rather than a single one, and the bracket syntax matters. Without the brackets it will not do what you expect.

https://cdn.filestackcontent.com/zip/[VK2LSKTSTmuS6U5fqFkp,jLoemkNMSiyxhKHxo4G0]

The photograph and the document come back as one archive of 2,254,095 bytes, which you can download as a zip right now. If you want the download button wired into a real interface, our walkthrough on generating a zip with javascript file upload and React has the component code.

9. Store a processed file as a new asset

Transformations are computed per request and not saved. When you want to keep one, add the store task and Filestack writes it as a new file.

https://cdn.filestackcontent.com/store=filename:cat-thumb.png/resize=width:300/FILE_HANDLE

The response is JSON rather than an image, and it contains the handle of the newly stored file. Chain store onto the end of a resize or a format conversion and you have a permanent optimised copy without a build pipeline. One warning worth taking seriously, and the reason this is the one example here with no live link: a store URL writes a new file every time it is requested, so never put one in an image tag on a page. Call it once, keep the handle it returns. Our guide to running an image upload service covers when to store a variant and when to keep generating it on the fly.

10. Crop an image into a circle

The last one is small and genuinely useful. Avatars are the obvious case.

https://cdn.filestackcontent.com/circle/resize=width:700/VK2LSKTSTmuS6U5fqFkp

 

Note the result comes back as a PNG at 687,889 bytes even though the source is smaller. That is not a bug. A circular crop needs transparent corners, and transparency needs a format that supports it, so the task hands back a PNG. Chain a resize in front of it, as this example does, and the weight stays sensible.

Where the free plan stops

The free plan covers everything above and a good deal more, and it is not the whole product. Most AI-powered capabilities that read and interpret file contents run on the higher plans. Optical character recognition, tagging, moderation, captioning and video conversion all live there. Facial detection is the exception and runs on the free plan. If your use case depends on another model understanding an image or document, the plan you need follows from the operation you call.

What to do next

Take the URL shape from the top of this article, drop your own key into it, and work down the list. Every one of the ten returns something in a browser tab, so a few minutes tells you how much of your use case the free plan already carries.

The two lines you will reach as the project grows are the AI-powered operations and the monthly volumes, and the free plan comparison shows which plan clears each of them. Before that, the next thing worth reading is how these single operations chain into a delivery pipeline, which our guide to file delivery workflows walks through end to end.