February 13th, 2026
We’re reaching out to let you know about an upcoming change to how we manage AI-generated content on our platform.
To ensure peak performance and a faster experience for all users, we are implementing a new 30-day retention policy.
Starting February 16, we will only host AI-generated material created within the last 30 days.
What this means for you: Automatic Cleanup: Any AI content older than 30 days will be permanently removed from our servers on a rolling basis.
Action Required: If you have important AI generations that you wish to keep, please download them before February 16.
We want to make sure you don't lose any of your favorite work, so take a moment this week to save your files locally.
Thanks for being part of our community!
Best regards,
September 8th, 2025
July 11th, 2025
Upload any image and re-generate the same image WITH your trained AI Character.
How to use the img-to-img tool?
👉 Upload your target image
👉 Select one of your trained AI Character
👉 Click on recreate the image
Never prompt ever again!
June 25th, 2025
Upload any image and obtain a hyper-realistic detailed prompt. 👀
How to use it?
Let's say you like an image on IG or pinterest
✅ You upload it in the prompt generator
✅ You obtain the EXACT prompt to generate that same image
Benefits
✅ Saves you a LOT of time (you don’t have to build your own prompts anymore)
✅ Generate better images ultra realistic
June 24th, 2025
It's so much better than KLING 1.6 😭
It's the SAME price as KLING 1.6 💸
Enjoy now the best image to video on the internet!
June 3rd, 2025
Photoshop is DEAD. Long life to FLUX.1 Kontext.
Flux Kontext is an absolute revolution in image editing and compositing.
You don’t need ANYTHING else than a prompt to edit your images.
We created a full blog article on how to use Flux.1 Kontext, read it here. 👇
May 22nd, 2025
✅ Improved results (consistency + AI video quality) 👀
✅ Improved speed (2x faster) ⚡
✅ Much cheaper (2x cheaper) 💰
LoRa Training Method 3.0 Upgrade
Before all the trainings could only go up to 2000 steps
Now you can train up to 4000 steps 🔥
✅ More consistency
✅ Less deformed body parts
✅ Higher realism
March 31st, 2025
You can launch AI influencer trainings with up to 50 images instead of 20 images previously.
If you have a good dataset for your training with quality images, you should not need to add more than 15-20 images. We increased this limit upon the request of certain users.
January 25th, 2025
New Feature
TODAY we launched the NEW version of our AI video model.
How to use the AI video?
🖼️ Upload an image
✍️ Write a short prompt of the desired action (you can also write a negative prompt).
📱 Choose the aspect ratio (16:9, 9:16, 1:1)
What can you expect from AI video generator?
🎥 Ultra realist movements and physics understanding.
👀 Your AI video will EXACTLY follow the prompt description.
🏃♂️ Wide variety and complex actions.
🧑🎨 Insane creativity
🔥 Let's see what you can do!!!!
January 17th, 2025
Improved
Unlimited AI image generation at the same time
You can generate unlimited AI pictures simultaneously (before it was limited at one generation at a time)
We also improved the AI image generation speed
Bug Fixed
Fixed a bug where all trainings wouldn't appear on certain devices
Fixed Gems balance not properly refreshing after launching an AI generation
A woman standing on a balcony facing the sea