Autopilot
Every night the studio picks Advargo topics, writes the copy with the local AI, and makes posters and reels on your GPU — ready to post by morning.
Voice-over, captions & music
Upload your music track under Brand details (below). It's automatically lowered while she speaks.
Electricity & cost
GPU power is measured live with nvidia-smi; CPU power is measured too when the processor exposes it. The rest is your estimate above.
Brand details used on every post
Make content now
Your own jobs from other tabs always run first; autopilot work waits its turn.
Ready to post
Fresh topics from the web
OpenAI searches the latest study-abroad news for Indian students: visa changes, scholarships, intakes, universities and courses. It returns ready-made topics with the content for each one: key points, real names and facts with source links. Your studio then writes the script and makes the reel or YouTube video on your GPU.
Takes about 30–90 seconds while it searches and reads sources.
OpenAI settings
Checking…
Only the research request goes to OpenAI. Images, video and voice stay on your server. The key is stored on the server and never shown again. Web-search requests are billed by OpenAI per search.
Topic ideas
One click: branded poster or carousel from an idea
The AI writes the headline and caption, your GPU makes the photo, and it's designed in the Advargo house style automatically.
Finished posters appear in Autopilot → Ready to post, with caption and hashtags.
Or design one yourself: 1. Generate the background
Describe the photo. You'll add the text in step 2, so it's always sharp and spelled right.
Pick an image to design
Click a finished image to open it in the poster editor below.
Your generated backgrounds will appear here.
Generate images
Runs on your RTX 3060 through ComfyUI.
Recent images
Click an image to view full size.
Create a branded reel
Fully automatic: the local AI writes the story, your GPU renders it, and it comes out with Advargo branding, an end card and an Instagram caption with hashtags. Brand details come from Autopilot → Brand details.
Photo-quality scenes from your GPU with smooth camera moves, animated text and transitions. Best for posting.
Your reels
They keep rendering on the server even if you close this page. Time, electricity and cost are shown for each.
Advanced: raw AI video clips (unbranded, for experiments)
Generate video clips
Wan 2.2 on your RTX 3060. Expect several minutes per clip — you can leave the page and come back.
Reel branding
Used when you click Make Advargo Reel on a finished clip: 1080×1920 with your caption, brand bar and an end card.
Recent videos & reels
Raw clips are 24 fps MP4. Branded reels are 1080×1920 and ready to upload to Instagram.
Install an update
Upload the advargo-ai-studio.zip you received. The server backs up the current version, installs the update, restarts and checks itself — and rolls back automatically if anything fails. Running jobs resume afterwards.
Temperatures & cooling
When the GPU or CPU reaches its pause temperature, the studio stops starting new work and lets it cool. It continues automatically at the resume temperature. If the GPU hits the emergency limit in the middle of a render, that step is stopped, the card cools, and the step is redone.
The RTX 3060 is designed to run up to the low 80s °C and slows itself down above that. The defaults (pause 83, resume 72, emergency 88) keep it out of that range.
Previous versions
The last 5 versions are kept. Rolling back uses the password above.
Studio version
No updates yet.
Create a YouTube video
Long-form 1920×1080 guide: chapters, female voice-over, photos with camera moves, infographic slides, chapter cards, thumbnail, subtitles (.srt) and an upload pack with title, description, chapter timestamps and tags.
Topic ideas (click one)
Takes roughly 45–90 minutes on your RTX 3060 for 10 minutes of video. You can pause it any time.
Your YouTube videos
Download the video, thumbnail, subtitles and upload pack. Each card shows time, electricity and cost.