AE GPT Discussion.
Ask questions, report issues, share prompts, and help other users. Keep it public, searchable, and simple.
Search once. Then ask clearly.
Use the search box to find approved public comments on this page. If you still need help, post below with the Adobe app, version, provider mode, selected model, and the exact issue.
8 responses
-
Welcome to AE GPT guys! Leave your questions or suggestions here and I’ll be sure to reply to those.
-
Hello,
I am currently using AE-GPT with Gemini models through Google Cloud / Vertex AI infrastructure.
At the moment, AE-GPT appears to use the older Google AI Studio Gemini endpoint:
`generativelanguage.googleapis.com`
This causes billing to go through direct Gemini API billing instead of Vertex AI billing and cloud credits.
I would like to request support for:
* Vertex AI Gemini endpoints
* Google Cloud API keys / Vertex authentication
* OpenAI-compatible Vertex endpoints
* Service account or Vertex API based authentication
* Ability to select custom Gemini base URLs/endpointsMy current issue is that Vertex AI API keys do not work inside AE-GPT because the app still targets the older Generative Language API endpoint.
Example desired support:
* Vertex AI Gemini 2.5 Flash / Pro
* `aiplatform.googleapis.com`
* OpenAI-compatible Vertex endpoint supportThis would allow users to:
* use Google Cloud credits properly
* manage enterprise billing
* avoid AI Studio billing limitations
* integrate with production cloud workflowsPlease let me know if Vertex AI support is planned or if there is already a workaround available.
Thank you.
-
Hi Vignesh, yes the AE GPT currently supports Gemini APIs through the Google AI Studio. And as of now Vortex AI authentication is not in works but I’ll do some research and see what are the possibilities and limitations.
Thank you.-
Thanks for the update, Umair.
I understand about Vertex AI not being in works yet. I appreciate you looking into the potential for it. Looking forward to your research results.
-
-
-
I’m not sure if its intelligence comes from the LLM or the project understanding method from the plugin’s skills. I’ve tested it and it hasn’t delivered anything useful yet. Can you point out some key benefits? For example, could you make a video tutorial demonstrating at least 10 practical benefits it offers? Thank you.
-
Yes, its intelligence comes from the AI model being used. For example a small local LLM can’t do much in comparison to GPT 5.5 or Opus 4.8.
This tool basically takes a prompt from the user, then optmizes that for After Effects using the AI model, and then that optmized prompt is used to generate Extendscript for the task user asked and this is where small models fail to produce a working Extendscript.
I’m working on a video to demonstrate its uses. One this I would like to mention, it’s not like “do this all for me” but rather it can help you speed up your workflow in tasks that take time.
-
-
Will there be a way in the future to access your gemini subscription in addition to calude and gpt chat to avoid using the api key?
-
Hi, sure. You can use your Gemini subscription through the Gemini CLI. It has been supported since last few versions already. You can even use your free Google account (with small quota for Gemini).
Though in my experience Gemini does not perform as well in the After Effects Extendscript thing as the the others from OpenAI and Anthropoc perform.
-
Leave a Reply