AI PaintingI can say that it's quite mature.
Closed-source model Midjourney, easy to use, can produce photography-grade work, the effect is awesome. It is a few dozen dollars per month, a little costly, and there is poor scalability.
who require a high level of consistency in character scenes, or who want to use a workflow as an exclusive drawing tool application.Open SourceIt's the models that are YYDS, like the stable diffusion series, and the Flux, which has been a huge hit lately, and the self-publishing platforms are being butchered by these two.
The charm of open source, lies in the scalability, controllability, but also with the workflow packaging products, and the current effect is constantly approaching Midjourney, the disadvantages are also very obvious, the learning threshold is very high.
How to play with these open source painting models for the little and average person?
There are currently three ways
1、Purchase high-computer and install the model by yourself, which is costly and troublesome.
2. Renting a cloud computing platform and then deploying the model is still more troublesome
The 3rd, which I'm going to share today, is to use cloud-based open source modelsPainting Platform, is also one of the easiest to sign up for and use online.
Tool name: tusiart (Chinese name:toast (loanword)）
address:https://www.1ai.net/1530.html
Just register and login with your cell phone number or WeChat.
Every day you can give away 100 arithmetic, text born map 1.2/picture, map born map 1.6/picture, equivalent to be able to free whoring 60 to 80 pictures, very suitable for newbies to enter the pit.
What can Toast AI do?
1、Online AI painting
Toast has gone online with the SD series of models,FLUX, Hybrid DiT, etc., free choice at the touch of a button.
Choose a good model, other parameters, you can generate pictures online, there is no requirement for computer power, everyone can run. There are rich plug-ins, model support, good effect, high playability.
Generate all kinds of beautiful pictures, rare and strange pictures, are small, the cap and scalability is unbeatable.
If you don't know how to write a painting cue word, it is recommended to find some cases to learn, it has a cue word parsing function above, upload a picture, you can parse to get the cue word of the reference picture.
2. Citing community fine-tuning models
Home model here, there are a lot of other people tuned model, like which, you can click a click to copy the model and parameters.
3. Establishment of ComfyUI image workflow
The workflow enables the creation of a variety of interesting AI graph-generating applications that can be used over and over again to improve graph-generating efficiency.
Also, we are able to view and copy other people's workflows in the workflow community.
4. Training models
For professional players, it is also possible to upload, or train your own models.
Finally, toast belongs to the open source model of the painting platform, can provide a simple way to get started with open source models, but to play to understand stable diffusion and other open source graphical models, ComfyUI workflow, etc., or need to draw a lot of time to study and learn.