- Hypothius News
- Posts
- The 2025 Ultimate AI Filmmaking Guide
The 2025 Ultimate AI Filmmaking Guide
Your guide to free and paid AI text-to-video software
🎬 Restaurant That Employs People With Disabilities: The Aregak Cafe
People with disabilities want to feel like the full members of society that they are.
This amazing restaurant employs people with disabilities that many companies refuse to hire.
It shows that they are of course actually able in so many ways, and that it is much of the world that is closing opportunities for them.
This highlights the importance of inclusion and acceptance in the workplace.
Hovo, the server in the above video, is a host, baker and barista. He has down syndrome, and lives most of his life independently.
Check out the above video that has 255,000+ likes.
As the video explains, the three primary goals of this cafe are to:
Train and employ young adults with disabilities and their family members
Challenge stereotypes surrounding disability by showing ability in action
Generate profit to support their partner therapy center for kids with disabilities, called Emili Aregak.
This is the future. This company is operating in a way that is ahead of its time.
I think all businesses can learn a little something from them.
➡️ Want to learn how to make a difference with your company?
Join the Hypothius Changemakers Community on Discord and connect with likeminded individuals working to create lasting change in the world.

Top Free AI Video Generation Software
Do you want to make ai videos on your own but you don’t know how, and you want to use something free? Here is a good place to start.
Pika - has some EXTREMELY impressive special effects options. Also, last I checked, you can create free unlimited video generations using their early model on their Discord channel. Pika just came out with an insane new feature (paid) called “scene ingredient”, where you can add visuals for characters, object and location, and boom - your custom special FX scene.
Vidu AI - really amazing with human movements, facial expression emotions, and allover cinematic framing, and impressive prompt adherence. Not just slow-motion animation - lots of movement! My personal favorite. It now has reference-to-video, meaning. Apply for their Video Artists program, where you can get free credits by creating and uploading edited ai videos onto your social media accounts - they’re one of the only video generation platforms that runs something like this. Vidu also allows for reference-to-video - “generate videos with multi-subject consistency across elements.” - basically the same as Pika’s “scene ingredients” feature - give it your props and location and character, give it direction, and you’re good to go.
Krea - awesome for dream-like creations in the “animator” timeline tool. Mouse over the “generate” button and select “animator”.
Luma Dream Machine - also impressive, but might be more prone to glitches such as extra arms and legs. What makes this one a bit more unique is that by. using their “brainstorm” feature, you can generate preview images of your videos before hitting the “generate” button, for deeper accuracy(Runway also has this feature, but its paid). It was also the first openly-available ai video generator that had so much consistency from frame to frame and didn’t mesh all the frames together into abstract transformations. It also has a beautiful interface. 30 videos a month.
Kling AI - unfortunately, I recently received a message in the free version saying that the estimated wait time was 3 hours for one generation. Make sure to use this platform during the hours of the day that are less busy…perhaps at night or not during the weekends. But it has very impressive attention to detail. Check out this cool generation of a fish bus, and this impressive generation of a female video game character. Very realistic subtle movements. Can generate up to 3 minute consistent video in the paid version, which is something that sets this model apart.
Haiper AI - many different unique ai tools
Hailuo (aka MiniMax) - I feel like this model is also really good with realistic human movement and facial expression emotions. Very impressive and realistic mouth movements for speech, although custom speech is not yet synchronized. High-quality consistency. Some extreme realism over here. Check out this crazy example video of their new model.
With these platforms, to the best of my knowledge, you get a certain amount of free credits, every day/month. If you use all of these platforms in one day, you will have a good amount of free credits per day. Also, if you have a computer with high VRAM, I highly suggest setting up ComfyUI and Stable Diffusion for the ultimate array of unlimited creative possibilities.
What are the most famous text-to-video generators in general, both paid and unpaid?
Here is the full list of both paid and unpaid industry standard ai video generators.
Hailuo (aka MiniMax)
Runway (recently had a collab with Tribeca Film Festival, and opened division at Lionsgate). Only a small amount of free credits when you first open your account.
Adobe Firefly (coming soon)
Google’s Veo 2 (coming soon) - check out these cinematic outcomes.
What is the difference between text-to-video and video-to-video?
Vid2vid means that you take a pre-existing video, and put it through AI to transform it. Runway, Stable Diffusion, and Haiper offer this service.
There is also image-to-video, which all of these platforms offer. Some platforms, like Kling, offer a motion brush, meaning you can paint in one direction you want different things in the image to move towards, for detailed control of the animation. How cool is that?
Paid All-In-One Alternative
Do you want to try and use all the platforms, and incorporate it into an online video generator? Then try Rubbrband Movies, the only platform that combines nearly all of these generators into a single easy-to-use interface.
How can I edit these videos and make a movie?
123Apps has a great quick browser-based video editor. I suggest stealing that workflow. For a more detailed editor with more polished content-creator tools, check out CapCut. If you want to edit your footage on a software that is downloaded onto your computer, I suggest using Avid Media Composer First.
What about music?
How can I learn more about AI?
Check out the website Curious Refuge for awesome contests, opportunities and more. Also check out the Runway Global Talent Network that was launched under two weeks ago - it’s a brand-new space to exhibit your work and find jobs.

Information about Luma’s new Dream Machine model via their original website.

The general look of an ai text-to-video generator. This one is from Vidu 1.5.

Some ai text-to-video generators have colorful splash pages with lots of features.
AI Video Generation and Special Effects
I am very excited about Adobe Firefly’s future capabilities with AI, special FX, and film! Check out the video below to see just what I mean.
@thepacman82 Adobe Firefly has generative video now?? Let’s take it for a compositing test drive. All video elements generated with @adobevideo Generat... See more
Pretty soon it’s going to be incredibly easy to mask things out with video in Adobe Premiere, using just the press of a button, no green screen required, thanks to AI. Expect the video scene to get a lot more artsy in the coming months.
New AI Video Trend: Templates - The New Special FX
Similar to the scene ingredient feature I explained about Pika, Templates are where you can put yourself into realistic scenes, making special effects that knock people’s socks off.
How can I do these special effects now?
You can use Pika’s features, Haiper’s features, or Vidu’s features for this. Or you can do it manually, like I did.
Let me break down the process for an AI commercial I created for a competition.
In Greek Mythology, Tethys is a Titaness and the embodiment of the nourishing flow of water. She is the mother of the rivers, fountains, and clouds. Tethys represents the nurturing aspect of water, providing sustenance to all life.
In this film, after drinking her Path water, a bored young woman wakes up and transforms into Tethys, and sets out to turn garbage into plants and rebuild war-torn cities.
The concept is that when a person drinks Path water, they can "flow like water".
We used CapCut for some of the transitions, and for the live-action cutouts, color correction and the filters, the clips were generated using Vidu, and we used Adobe Premiere for the final cut.
For the water-woman transformation, I screenshotted the first frame of the shot of me, put it through Adobe Firefly to match the general structure of the image while turning it into an office, then I brought it into adobe express and erased the figure from it and added in a still cutout image of myself from that first frame, after that I brought it into Vidu and animated it, then I matched that first frame with the frame exactly before it in the real video so that it would transition into a transformation and I added a white flash to help with transforming.
I did that technique for some of the other shots too, and tracked the blur over the water bottle (that mixed with black and white helped hide the fact that it wasn't the actual bottle), and I think the black and white/distortion actually started making it difficult to detect where the real footage began and the ai ended, which I thought was pretty cool.
I also tracked myself on the moving ai shots at the desk so that it wouldn't look like the background was moving and I wasn't. But I made sure to shoot at a real desk so that the AI desk/office would be a bit more realistic the way I leaned on it.
The music was AI and made with Udio, and for the closeup shot, I took the original shot from Vidu, used a face-swapper from Hugging Face spaces, and then brought it back into Vidu for image2video.