Omni-AI: Combine ChatGPT with other AIs

(Limitation: Support picture (experimental: video) creations (excluding modifications) only at current). Changelogs

https://github.com/Huilin40770207/combine_chatgpt_with_other_ai

OpenAI API Key (see the link for help):

Replicate API Key (see the link for help):

What do you want to do:

(Optional) Input URLs:

Examples of queries:
(1) Generate a paragraph of a beautiful Chinese female human with four sentences only. After that, generate a photo of the female based on the paragraph.
(2) Describe a beautiful female with four sentences. After that, generate a photo of the female based on the text.
(3) I'd like to furnish a bedroom. Please suggest me some popular furniture. Then show me one photo of the room furnished with these furniture.
(4) Generate a story of two cartoon characters Tom and Jerry. Then summarize the story into four sentences. Then generate a photo of Tom and Jerry based on the summarization.

(New-experimental) Generate a video of "a panda walking on a street"
(New-experimental) Modify images and replace the room to be "a 2020-style bedroom for kids"

Limitations:
(1) SLOW. It takes ~40s to generate a story and one picture.
(2) BUGGY. I'll improve the backend with my best effort!

Comparison with Other AIs: