Free to use AI image generators.
Free to use AI image generators.
Author
Discussion

2172cc

Original Poster:

1,767 posts

122 months

Tuesday 17th March
quotequote all
I've been using the basic Google Gemini for about 8 months and have generally been pleased with the results, using it mainly for creative imagery but recently it's been giving me some odd results and the 'only a language model' message. While I find it very easy and straightforward to use it's starting to become unreliable.
I tried the basic version of Chat GPT yesterday but it only gives a relatively small daily limit meaning I have to wait till tonight before I can use it again although it did work well and its very intuitive.
A quick search threw up Freepik and I tried that but it's not very user friendly and a bit too complex for me unless I spend some time going through the tutorials.
What recommendations are there for free easy to use AI image generators that are straightforward and have a decent daily limit?

GuyW

1,115 posts

228 months

Tuesday 17th March
quotequote all
Where's the you can pick two out of the three triangle diagrams... biggrin

I think you'll either need to invest a little time to get used to some of the more obscure interfaces and prompts to get what you need or look to invest in a non free option.

Prak

837 posts

243 months

Friday 20th March
quotequote all
I try not to use AI too much but found that duck.ai gives some surprisingly good and quick results without any login. Not sure what daily limits apply but if has no login then separate private browsing windows might let you keep going for a while?



prompt said:
A man dominating the stairs while holding a tin of custard powder in one hand and Red Bull in the other. He looks angry.

Mr Whippy

32,423 posts

266 months

Monday 23rd March
quotequote all
If you have a decent gpu and a chunk of hdd space try comfyUI.

Free yourself from the limits of prompting and single stage approaches and go full creative mode by mixing all kinds of inputs etc.

captain_cynic

16,469 posts

120 months

Monday 23rd March
quotequote all
Mr Whippy said:
If you have a decent gpu and a chunk of hdd space try comfyUI.

Free yourself from the limits of prompting and single stage approaches and go full creative mode by mixing all kinds of inputs etc.
This, if you've got a decent GPU it'll be faster than the online versions and better quality as you can produce higher res images/upscale.

Biggest issue is getting the right resources for what you want (checkpoints, loras, et al.). What the AI companies are selling isn't the tech, rather the training data.

2172cc

Original Poster:

1,767 posts

122 months

Monday 23rd March
quotequote all
thanks for the advice

AlexC1981

5,619 posts

242 months

Monday 23rd March
quotequote all
captain_cynic said:
This, if you've got a decent GPU it'll be faster than the online versions and better quality as you can produce higher res images/upscale.

Biggest issue is getting the right resources for what you want (checkpoints, loras, et al.). What the AI companies are selling isn't the tech, rather the training data.
I thought my 5070 TI was a "decent" GPU until I looked up the requirements eek