brave new launch
from text… straight to mind-blowing images and videos without overthinking prompts. any format. any language. simple ui. simple api. and no forced subscriptions you forget to cancel.
this is the promise
“maginary - invoke the unseen” at 10% volume goes well with this post
what was mjapi again?
an image and video generator that reads and blows your mind
so what
this is the more official launch day, call it v1.0 if you wish
the soft-launch from a few weeks ago (damn, it’s actually 2 months already? time flies) was just my attempt to market this thing. it went pretty well
results
so it feels good to get some traction, things are going well. could be better, but I see enough validation to move forward
-19_09_2025.png)
yeah, I messed something up around sep9, since sales were constant and then suddenly stopped until a few days ago. thing is, there’s no clear sign. could be some ui bug, or any change in my yapping about the project on x/reddit. who knows
btw shout-out to @metabase for the amazing bi graphs, I’m using their free self-hosted docker image and it works great!
what mjapi does now
it does text to image, text to video, image to video. but let me show you what that actually means
majestic old tree in a fantastic setting full of life
then upscale it, vary it, zoom out, pan (add pixels), or turn it into a video. all from the same starting point, here are the current available controls for most formats:
for example, here’s a zoom out:
“hmm, I wonder what’s in the tree?”, just pan the camera bro:
need a logo? works with any format, including vector graphics
logo for cat food “kitty kit” --svg
want clean line art?
beautiful tiger line art, monochrome, simple lines
or get the style from any other image:
portrait of a woman holding a cup of tea --sref https://s.mjapi.io/styles/line_art_4.png
or combine your own images with style references:
https://s.mjapi.io/demo/man-in-coat.jpg --sref https://s.mjapi.io/demo/impressionist-style.jpg also add a cigar
and yes, you can write prompts in any language. they’re translated to english internally:
un chat noir élégant dans un jardin japonais
videos baby
this is big. text to video, image to video - it just works
pink frog jumps --mp4
take a photo of you directly via the website (or upload one). its url gets generated/pasted instantly in the prompt (before it’s uploaded. yeah, magic baby)*
https://s.mjapi.io/u/8b2e24a0c6850dc0e11611087d6d2260.webp make me blonde and get me a badass beard
app.mjapi.io/g/b83a8784c4081b1ad70c81a29b473c89
then the video ▶ button and describe it:
this guy stands up and leaves the room --720p --5sec
main parameters
here’s what you can control if the UI isn’t enough. and this list will obviously grow over time.
aspect ratios, style references, model selection, output formats — it’s all there.
what changed since july
- video generation: complete implementation with multiple cutting-edge models
- style references (--sref): reference specific visual styles in your prompts
- progress tracking: real-time generation updates so you know what’s happening
- credit system overhaul: new pricing tiers (no-subs: novice; subs: acolyte, mage, archmage)
- generation history: see everything you’ve created on your homepage
- api access: proper api keys and documentation for developers
- image upload: reference your own images with frontend preprocessing
- chill audio player: because waiting for generations should be pleasant
- image picking: select and focus on specific results with smooth animations
- mobile experience: comprehensive UI improvements, responsive everything
- some infrastructure scaling: added more celery workers, parallel processing of each of the 4 slots, redis caching
- probably some other important stuff I can’t remember rn
mjapi evolved from just standard image gen to a comprehensive prompt-based creative tool
and, as I was saying, you can listen to some cool chill original soundtracks while creating. just scroll to the bottom of the page and you’ll see mjapi logo turning into a vinyl:
yeah, totally unnecessary, but given that I “designed” the logo without even thinking about its resemblance to a vinyl, when I realized this I HAD to have it. and here’s the cute player:
and yeah, if you’re seeing those 2 icons there, you already know what happens next.. you’ll get the album on apple music and spotify. because, why the heck not? it’s just 15 bucks with cd baby
tagline change!
former midjourney api, now something better -> an image and video generator that reads and blows your mind
try it
- app.mjapi.io - no account needed to try the sample prompts!
- mobile app coming soon
the build in public part
I’ve been building this in public, sharing the ups and downs and iterating on your feedback. so you, as an early adopter, deserve something special: here’s a lifetime 10% off code: SOEARLY10 it’s valid until oct1 ‘25 for 100 users
btw, here’s my users per country: 
what’s next
mjapi is not trying to be the next midjourney. just something different for developers and creators who want:
- beautiful and diverse outputs by default, without overprompting
- video generation alongside images
- reliable apis that just work
- reasonable pricing without surprise costs or forcing you to subscribe to use the app
- a platform that evolves based on actual user needs
this is the foundation. we, together, will build on top of it through your feedback, so please share your experience! I see very few users actually email/dm me about issues/feature. guys! I’m here to build whatever the heck you need, for ‘free’! when we’ll reach 1k users you won’t have that luxury anymore (unless your demands are popular)
next is scaling. let’s do it. 
create: app.maginary.io
follow my progress: @xucian_
hang out: discord server