Summary of "구글 Gems 미쳤다... 나만의 Gems 2개로 AI 애니메이션 만들기 (프롬프트 공개)"

Overview

The video demonstrates a fast, repeatable workflow to make short AI animations using two Google-based tools called Jams (an image-generation Jam and a video-conversion Jam). The workflow is powered by a Gemini-like chatbot and Google’s Opal automation. It’s presented as free to use and targeted at beginners and intermediate creators to reduce repetitive prompt/restore steps.

Main workflow (two Jams)

1) Image Jam — generate frames

Purpose

Produce a single 16:9 composite image made of a 3×3 grid (nine equally sized frames) to act as animation frames. Nine frames can represent ~40–50 seconds of video potential.

Creation

In Jams → New Jam, supply name, description, and a “request” prompt.
Instead of hand-writing complex prompts, use the Gemini chatbot (New Chat) to auto-generate the detailed prompt.

Key prompt requirements (used by the presenter)

3 columns × 3 rows within a 16:9 image (nine frames).
Thin colored borders (different colors) around each cut so frames can be individually identified and extracted.
Preserve the uploaded subject’s exact breed/fur/face/body shape (useful when uploading a pet/character photo).
Scene template variables limited to three inputs: play surface, location, character.
Cut plan: shots 1–3 close-ups (ingredient prep), 4–6 cooking process, 7–9 finished dish + character reaction.

Examples and notes

Examples used: uploaded puppy/dog or a Pixar-style 3D character.
Test scene: a wood-toned kitchen and a pizza dish.
Model mentioned: “Nanobana 2” in fast mode (model name referenced in the transcript).
Save the Jam; use “Create Image” to generate composite images.

2) Frame extraction

Use the colored thin borders to extract individual frames by prompting the Jam (example prompts: “Extractor Blue Frame”).
To remove the colored border, follow up with a prompt like “remove the blue line” to yield a clean frame.
Download extracted frames. The presenter provides exact English prompts and alternative prompts in the pinned comment for failure cases.

3) Video Jam — convert frames into short animated clips using Opal

Process

Use New Jam (English) which runs an Opal automation app (referred to as “Frame Animator” or “Untitled Opal App”).
Upload a frame image into the Opal prompt window; Opal interprets the image and generates a short, natural-flowing video clip. Opal typically adds situational sound effects but not background music.

Opal advanced editor settings (steps to set)

Open Advanced Editor → select the single “Video” step → set:
- Agent: VO
- Advanced Settings: VO 3.1
- Aspect Ratio: 16:9

Workflow notes

Start the automation, upload frames in order, and let Opal generate clips.
The system may auto-adjust sequence/order and sometimes insert unwanted actions (for example, an extra door or pizza tray appearing). These can be excluded or trimmed.
Final assembly: import generated clips into an editor (CapCut, Premiere Pro, etc.) for sequencing, trimming, and final editing.

Practical tips & cautions

Automate prompt generation with Gemini to avoid repetitive manual prompt writing.
Use colored thin borders to easily identify and extract frames; use targeted extractor prompts for each frame.
The system auto-handles many steps; you mainly supply images and minimal prompts.
The presenter used multiple “pro accounts” to speed testing.
The workflow is presented as free, but avoid abusing it to farm low-effort, auto-generated videos.

Warning: blindly churning out auto-generated content can risk YouTube policy problems or channel removal. Use the workflow to increase efficiency and add creative planning, not to bypass content-quality expectations.

Resources and support mentioned

The creator will post all prompts used (English prompts and alternatives) in the pinned comment for step-by-step reproduction.
Plans mentioned for offline/online coaching classes and a community “cafe” for group learning; advance announcements will be in the pinned comment.

Tools, models and apps referenced

Google Jams (image Jam and Opal-based video Jam)
Gemini / “Jeminai” chatbot for prompt generation
Opal automation (Frame Animator / Opal App)
“Nanobana 2” (image-generation model referenced)
Video editors: CapCut, Adobe Premiere Pro

Main speakers / sources

Video creator / presenter (channel owner — referenced as “Hadang” in relation to pinned comments)
Google tools: Gemini-like chatbot (Jeminai/Gems), Google Jams, Google Opal
Models/tools mentioned: Nanobana 2, CapCut, Premiere Pro