What's New?
1- AI Model Comparison
We have developed an interface that allows you to dynamically test multiple AI models. Through this interface, you can run multiple AI models using a single prompt.
2- Each IMG Upscaler
We received a lot of requests for an IMG Upscaler, and we've released the first version of Each IMG Upscaler. You can directly add it to your workflows to enhance the photos you produce.
3- Client Cache
Until now, most of our improvements have focused on the API and SDK. After receiving feedback, we're placing more emphasis on the UI. As a first step, we've enabled client caching.
4- AI Models
New AI models have been added to the platform.
* ms-img2vid
* stable-diffusion-3.5-medium
* stable-diffusion-3.5-large-turbo
* recraft-v3-svg
* recraft-v3
* florence-2-large
* omni-zero
* controlnet-1.1-x-realistic-vision-v2.0
* pixart-xl-2
* sdxl-controlnet-lora
* sadtalker
* moondreamer2
* trip0-sr
* stable-audio-open-1.0
* depth-anything
* lcm-animation
* stable-diffusion-inpainting
* sdxl-controlnet
* video-crafter
* illusion-diffusion-hq
* ideogram-v2
* ideogram-v2-turbo
* stable-diffusion-3.5-large
* hailuo-video-0
* hailuo-video
* pyramid-flow
* image-compressor
* each-upscaler
* nsfw_image_detection
* flux-dev-controlnet
* frame-interpolation
* hailuo-video
* realistic-vision-v3-inpainting
5- Ready to use Workflows
New workflowshave been added to the platform.
* Expectant Parents Generation + Pika
* Product Commercial Maker (Hailuo)
* Lego Character Generator
* Podium Video Generator (Hailuo)
* Halloween Lipsync Generator
* Halloween Image Generator
* Halloween Video Generator (Hailuo)
* Pika Art Community Edition
* Expectant Parents + Hug (Hailuo)
* 3D Cartoon Video Generator
* Photo to Anime Generator + lipsync
* Short Film Generator
Bug Fixes
1- The issue where the Output tab continued to display the previous flow's output after triggering a flow and switching to another one has been fixed.
2- The UI distortion issue that occurred after a flow was locked has been fixed.
3- The issue where the page would extend downward if the model's prompt field was long on the Flow page has been fixed.
4- The bug that prevented error messages from being closed in the UI has been fixed.
5- The bug that caused errors when selecting and providing preset inputs to models has been fixed.
6- The issue encountered during platform registration has been fixed.
7- The UI distortion issue that occurred when adding or removing a model on the Flow page has been fixed.
Improvements
1- Flow Architecture Update
We made a significant and challenging infrastructure change to ensure flows run faster. With this update, flows now operate 300% faster!
2- Designs Ready
The designs for Settings (Billing, Team, Audit Logs) are ready. We plan to activate them next month.
Partner Integrations
1- Hailuo AI-Minimax
Minimax has developed large-scale Mixture of Experts (MoE) models using multiple modes. Additionally, is the creator of Hailuo AI and it offers tools for creating high-resolution video clips through the MiniMax AI platform.
2- Beam Cloud
Beam Cloud is a fast and affordable platform that offers a user-friendly Python interface for managing serverless workloads. Users can utilize large amounts of GPU by automatically scaling their virtual servers and pay only for what they use.
