I’m an indie dev building FreyaVideo, an AI video hub that connects multiple models into one simple interface.
Instead of trying to build “the final product” from day one, I’m running one small experiment per week. This post is a quick write-up of what I’m doing and why.
The problem I see
More and more people want to use AI to create video:
- Product demo videos
- Promo / ad videos
- Short social clips
- Simple explainers
But in reality, a few things are painful:
- There are too many tools and models.
- Each tool has its own UI, credits, limits, and rules.
- If you’re a solo founder or small team, you don’t have time to learn 5–10 different products.
Often, you just want:
“Give me one place where I can type my idea, pick a model, and get a decent video out.”
That’s the starting point for FreyaVideo.
What FreyaVideo is (right now)
FreyaVideo is not trying to invent a new model.
Right now it’s a hub that connects three models:
- Veo 3
- Sora 2
- Nano Banana Pro
You write a prompt (or use a preset), FreyaVideo sends the job to one of these models, and gives you back the result. You can pick a model based on:
- Speed
- Quality
- Budget
The current focus is simple:
- Make it easier for indie hackers, solo founders, and small teams
- To use strong video models
- Without learning multiple dashboards and payment systems
Over time, I plan to:
- Add more models
- Add more very specific use cases / scenes (e.g. product demo, app promo, feature teaser, etc.)
- Let users choose a “scene” first, then a model
You can see the current version here:
👉 https://freyavideo.com
How I’m building it: one small experiment per week
I don’t have a big team or a big budget, so I’m using a simple rule:
Every week = one small experiment.
An “experiment” for me usually looks like:
- Ship one focused page or feature
- Drive a bit of traffic to it
- See what happens (clicks, signups, questions, confusion)
- Keep what works, delete what doesn’t
Examples of weekly experiments:
- A new landing page for one specific use case
- e.g. “AI product demo videos”, “AI app promo videos”, etc.
- A small UX change in the flow (how you pick a model)
- A different way to explain pricing and value
The important part: I decide the experiment, ship it, and then force myself to review the numbers at the end of the week.
This week’s experiment
This week, my experiment is:
Make it easier for people to understand one clear use case,
instead of trying to explain everything at once.
So I’m:
- Focusing the copy around a single scenario (for example: product demo video)
- Cleaning up the page so the flow is:
- Who this is for
- What problem it solves
- What you get (example)
- One clear call-to-action
And then I’m sharing the page in a few places (like this post) to collect:
- Click data
- Time on page
- Feedback / questions
The live version of FreyaVideo is here:
👉 https://freyavideo.com
What’s next
Short term:
- Add more focused pages for different scenes (product demos, promos, app teasers…)
- Improve the flow of choosing between Veo 3, Sora 2, and Nano Banana Pro
- Make it clearer when each model is a better fit
Long term:
- Support more models as they become available
- Let users pick a scene first, and then auto-suggest the best model
- Share weekly experiment results in public (what worked, what failed)
What I’d love feedback on
If you check out the site, I’d love to hear:
- Is it clear what FreyaVideo does?
- Would you ever use something like this for your own projects?
- What’s the biggest thing missing before you’d trust it with real work?
You can comment here or reach out to me directly — I’m happy to share more details, numbers, and future experiments as I go.
Thanks for reading. 🙌
Top comments (0)