I cancelled IMS and frame.io. Here is how I slice, tag and sell stock footage with one app (Clip Up Pup tutorial)

If you upload stock photos or videos and hope to make a bit of passive income, this is for you. I did it the slow way for years. I did it the expensive way for a couple more. Then I got fed up and built my own app. It’s called Clip Up Pup, it’s a one-time purchase, and the full walkthrough is below. Everything I say in the video is written up here too, so you can skim, search it, or listen to it while you make a coffee.

The short version: Clip Up Pup takes one exported timeline from Final Cut, Premiere or DaVinci Resolve, slices it into individual clips at your cut points, writes the title, keywords and description for every clip with AI, bakes that metadata into the files themselves, and sends them straight to the stock agencies. The AI runs on your own Mac for free. You pay once. There is no subscription, and there never will be.

Rather listen? The whole write-up as audio, about 17 minutes.

Read aloud by Ezay Read, the little Mac app I built that reads any text you highlight, on your own machine, with nothing sent to the cloud. If you would rather listen than read, it is yours for a one-off price too. No subscription. You may be noticing a theme.

How stock contributors actually get this done

There’s a few different ways you might do it. One is that you go to each agency one at a time, fill in the metadata yourself, and press upload. For every single agency. Maybe you do it another way, which is you go to ChatGPT, explain the photograph or the video, ask it to write the metadata, and then copy and paste that over to the numerous different agencies you use. Or, if you were like me, you use something like IMS or Stock Submitter or Microstock Plus, and they use AI to help tag your photographs and videos before they get uploaded.

I used IMS, and I thought it was the bee’s knees when it worked. It saved me, honestly, years of time, and I’ve made videos singing its praises. But the other half of this story is that if I was uploading videos, I’d have to cut them up and add them all individually. Until I found frame.io, which offered a service that created clips from one timeline. You’d export one timeline and end up with 200 different files that you could then run through IMS. I started using both of those and saw huge results across different agencies. I still generate passive income from those assets today.

Why I stopped paying for IMS and frame.io

There were a few problems. frame.io alone was incredibly expensive. At the time it was costing me about £140 a year, and it has plenty of other uses, but slicing up stock footage is primarily why I bought it. I use Final Cut, and it doesn’t offer a very good solution for exporting lots of different video clips at once. Even now it has a feature that sort of does it, but it still isn’t quite what I’m after.

And then IMS. I ended up paying for a bigger subscription because I had so many assets going through it, and that was costing me a lot of money every single month. The annoying thing was that when I was running it, it was taking up a lot of memory. My computer was working incredibly hard to think of all those thumbnails, titles, descriptions and keywords, and I was paying a subscription to use an outsourced AI. That just didn’t sit with me.

So after about six months of frustration with both, pricing and technical difficulties, I decided to make my own. I spent over two months on it. And I didn’t just want it to do tagging for me. I wanted it to slice up a timeline. I didn’t just want it to make clips, I wanted it to embed the metadata within the files, so that when I store them on my computer that metadata is saved in the actual file, not just on a CSV. And I wanted to press a button that would send everything to multiple agencies so that when I logged in, I could submit very quickly.

Most of all I didn’t want to pay a subscription. That was the biggest thing. I had to hit a certain quota with my stock footage and photos before I made a profit, and that really bothered me. I worked hard to sell these assets and I’d have to sell a hundred of them before I was in the black. That sucks. Now I’m subscription free, and I want you to be subscription free as well. Clip Up Pup has no subscriptions for life, guaranteed.

What Clip Up Pup does

The idea is simple. You export a timeline from Final Cut, Premiere Pro or DaVinci Resolve as one piece of footage, along with its XML. You run both through the app. It slices the footage at your cut points, frame-accurately. It then assigns metadata using AI and bakes it into the footage itself. And it’s ready to ship to the different agencies, straight from the app. Slice, tag, embed, ship.

If you don’t need the slicing bit, because you’ve already got your footage cut up or you’ve got another way of doing it, that’s absolutely fine. This isn’t really a slicing app at heart. It’s an adding-metadata app. The slicing was a very nice feature that saved me a subscription, so I added it in. If you just want to tag your photos and videos, drag and drop them into the app and go. Photos, videos, or a mix of both.

Why it’s free to run: local AI, on your own Mac

The reason I’ve been able to make it subscription free is this. With a little digging, what I found is that the companies I mentioned earlier use cloud AI models to tag and describe your videos and photos, and you pay for that every month. With Clip Up Pup, you download a local model and plug it into the application, and it’s free for life.

The models I use are called local, open-source models. I’ve been using them for about six months and they’re absolutely incredible. It’s a well-trodden path, it’s super well tested, and it’s private. I’m not going to ask you to download some dodgy malware. Go and check out Ollama for yourself; don’t take my word for it. Ollama is just the car. The model is the engine. The one I recommend, and the one that works really well for me, is gemma3:12b. It’s an 8 GB download and wants a Mac with 24 GB or more of RAM. Less than that, and the app will suggest gemma3:4b instead, which is 3.3 GB.

It’s totally local. This will never get back to me. I will never be able to see what you’re creating, and nothing leaves your machine unless you publish it. And the app itself is signed and notarised by Apple. I had to become an Apple developer to make that happen, and it’s all signed off.

Prefer to bring your own keys?

If you don’t like the idea of a local model, that’s absolutely fine. You can plug in an API key to something like OpenRouter, Anthropic or xAI and get them to do it via the app. I like to call all of my software bring-your-own-keys software. If you don’t know what an API key is, it’s like a digital key code that lets one piece of software talk to a different website. For this kind of task you really don’t need anything expensive. On OpenRouter’s Good tier you can get a great title, description and tags for roughly a penny per clip. The Great tier is about six pennies. You test the connection, you bosh it in, job’s a good ‘un.

Buying and installing it

You’ll find the app on the toolkit page, or go straight to the Clip Up Pup page. Buying takes you to Stripe, which is another layer of measures to keep your details safe. Once you’re through, you land on Your Apps, where your download and your licence key live. The key gets emailed to you too, so check your spam if you can’t see it. Any problems at all, email hello@fixitinpost.media and I’ll help you out straight away.

Drag the app into Applications. The first time you open it, macOS will ask if you’re sure, because it was downloaded from the internet. Yes, please. You’ll land on the explainer section, where every step has a little animation if you hover over it. I’ve put some fun titles in there like Sniff and Fetch. Then it asks you to choose your AI: local or bring your own key. If Ollama is running, the app detects it, picks up your model, and tells you the AI is ready. Paste your licence key in via Enter License Key in the top bar, and you’re in.

Tagging a folder of photos

Let’s start with photos. Pick your assets, give the shoot a name, put in the creator and the copyright. That copyright gets baked into your files, which I just think is a lovely feature. Then comes the bit that makes the difference: context for the AI. If you’re shooting a local area, like I do for my own stock website, you tell it. This is a shoot for the tourist board on the Isle of Wight during midsummer using a classic car on the coastline. Now the AI understands what the shoot actually contained, and every clip benefits.

Underneath that are stock profiles. Every export gives you one CSV per profile, and each profile can reframe the metadata. Adobe Stock gets the plain version. A site like Lightstock, which is a Christian stock marketplace, might want people playing in a park described as fellowship, discipleship, small group. Same shoot, different language, one tick box.

Press Read the folder, and the pup goes through all of your files. Then Draft with AI, and you watch it work. When I first built this I absolutely loved watching the metadata get added. A classic green MG roadster parked on a coastal path on the Isle of Wight. How cool is that? Keywords, description, category, all done. You can edit anything, add a product model if you’re shooting for a client, and there’s an Identifiable People flag that tells you which assets need a model release. That goes into the metadata and into a release-check CSV too.

When you’re happy, Embed in files. There’s a little progress bar as it writes everything into the actual files. Open one up afterwards, Get Info, and the keywords, title, headline and copyright are right there in the file itself. Drag those into Adobe Stock and it reads it all automatically. No fiddling with CSVs.

Night Shift: let it cook overnight

Because my AI is local and running on my computer, it can drain the RAM and make my Mac a bit sluggish while it works. So what I like to do is use the Night Shift mode. You queue up a whole bunch of shoots, pick when it should start, arm it, and go to bed. If your electricity is cheaper at a certain time, do it then. By the time you wake up you’ve got rich metadata on every asset you assigned. That’s what I do just before I finish work: line up a bunch of shoots that need metadata and let it cook overnight.

Slicing a timeline from Final Cut Pro

This is the bit that replaced frame.io for me. In Final Cut I lay out all my clips on one timeline, slow them down where I can to get more time out of them (they need to be at least five seconds), put a base grade on, and export the whole thing as one graded master. Mine are usually over an hour long. Then I export the XML of that timeline: Share, Export XML. It’s different for each editor but there are plenty of videos on it.

In Clip Up Pup, choose your graded master, attach the XML, and pick an output folder. One tip from painful experience: make sure that folder has enough free space, because if it runs out mid-export, it fails. Give the shoot a name, pick your format (H.264 is kind to your disk, ProRes if you’ve got the space), add the context, tick your profiles, and press Sniff the timeline. Lining up the scissors, reading the timeline, and there it is: every cut detected, the frame rate, the duration of each, with a filmstrip preview so you can check they’re all groovy.

Draft with AI, and it literally watches each clip. A couple holding hands on a grassy hill, framed by a doorway, overlooking a bright landscape. With the context I’d given it, that’s absolutely fantastic. Export clips, watch the counter in the bottom left, and you end up with a folder of sliced, tagged, embedded clips and the CSVs to match. From one shoot I’ll get 200 or 300 sellable files without cutting a single one by hand.

Sending to Adobe Stock, Shutterstock and the rest

Then you send them. You can pick an agency, add a destination, and as long as you’ve got the right credentials (the host, the port, the FTP details, which are on each agency’s site), you test the connection and send to all. Or, if you’d rather drag and drop, you drag and drop. Because the metadata is in the files, Adobe Stock picks up the title, the description and the tags on its own. Attach your releases if you’ve got them and submit. It’s as simple as that.

There’s also an export history in settings, because when you’re doing this alongside everything else it’s easy to forget where a batch went. And if you double-click the pup at the top, there’s a game of Tetris in there while you wait. I didn’t need to add it. I just thought it was fun.

What it’s not good at (yet)

You’ve heard me talk a lot about what Clip Up Pup is really good at, so here’s the honest bit. On IMS, the organisation side is really, really nice, and the connections with some of the agencies are slightly easier. Those are things we can work towards. But there’s lots this app does that IMS doesn’t: the slicing, your own local model, the shoot context, Night Shift. If you want to help me make the best possible product, give me feedback. The email address is hello@fixitinpost.media and every message gets read by me.

Client portals: what’s coming next

At the bottom of the export screen it says Coming Soon: Client Portals. If you’re sick of paying for Dropbox and Google Drive and frame.io just to deliver work, this is what I’ve been building. Portals where the client pays you for access, on your own domain, with streaming, instant download, their own desktop and iOS app, and if you’ve got Ollama plugged in it’ll even tag what they upload. I make more passive income from my client portals than I do from Adobe Stock most months. Clip Up Pup will publish straight into them. Join the client portals waiting list and I’ll email you when they’re ready.

Get it

You can buy Clip Up Pup on the Clip Up Pup page. The first 50 people to watch the video get it free for life with the code in the video description; after that there’s a discount code there for you as a thank-you for watching. If you’re a creator and you’d like to pass it on to your friends for a commission, there’s an affiliate programme waiting list too.

I’m not a big business. I’m just one guy who wants to save you money by cutting a few of your subscriptions. You’re supporting a solo creator who rolled the dice on a little startup, and I appreciate it.

Frequently asked questions

Is Clip Up Pup really a one-time purchase?

Yes. You pay once and it’s yours. There’s no subscription and there never will be. Updates to the version you own are included.

Do I have to pay for the AI as well?

No. The recommended setup runs a free, open-source model (gemma3 via Ollama) on your own Mac. If you’d rather use a cloud model, you can bring your own OpenRouter, Anthropic or xAI key and pay per clip, typically about a penny.

Which editors does it work with?

Slicing works from an exported master plus its XML from Final Cut Pro or DaVinci Resolve. Tagging works with any folder of photos or videos, from any editor or none.

Which stock agencies does it support?

It writes Adobe Stock and Shutterstock CSVs, embeds the metadata in the files so agencies that read IPTC pick it up automatically, and can upload over FTP to Adobe Stock, Shutterstock, Getty, Pond5 or any FTP server.

Does anything get uploaded to Fix It In Post?

No. With the local model, nothing leaves your machine. I can’t see your footage or your metadata.

Is it safe to install?

It’s signed and notarised by Apple under my developer ID, so macOS opens it without warnings, and everything the app needs is inside. No Homebrew, no Terminal.

What Mac do I need?

An Apple silicon Mac. For the 12B model, 24 GB of RAM or more is comfortable. With less, use the 4B model or a cloud key.