🎨 Meta AI
Generate AI images & videos with Meta AI (vibes.ai) — 4 modes: Text-to-Image, Text-to-Video, Image-to-Image, Image-to-Video. PLUS MAX
Overview
Meta AI is the AI content module running on Meta's vibes.ai platform (the code/backend keeps the internal name "vibes"). The page hosts 4 tabs:
| Tab | Name | Description |
|---|---|---|
| 1. T2I | Text to Image | Generate a 1K image from a text prompt — no reference images needed |
| 2. T2V | Text to Video | Generate 480p/720p videos from a text prompt |
| 3. I2I | Image to Image | Generate images from components: Character / Scene / Style |
| 4. I2V | Image to Video | Generate video from start/end frames or composed component images |
All 4 Meta AI tabs require a PLUS or MAX plan. Basic-tier accounts can browse but not run — the plan-required dialog pops up on Run / Run Selected / Retry Errors.
Meta AI uses a separate Meta account (the vibes.ai
meta_session cookie) — not shared with your Google/Gemini or Grok accounts. Aspect ratios: 16:9, 9:16, 1:1.Meta Account Connection
Go to Settings → Meta Account to add an account. Two methods:
- 🌐 Browser login (recommended) — the app opens a browser to sign in to Meta; the cookie lasts ~30 days
- 🍪 Manual cookie — paste the vibes.ai cookie (must contain
meta_session). You can use the Cookie Exporter extension to export cookies from vibes.ai
Add multiple Meta accounts to run faster (account round-robin). Enable Auto-disable on quota to skip exhausted accounts. Cookies can expire and require re-login.
Shared Configuration
| Field | Applies To | Description |
|---|---|---|
| Aspect Ratio | All 4 tabs | 16:9, 9:16, 1:1 |
| Threads | All | Tier cap: Basic 1, Plus 10, Max unlimited |
| Delay | All | Min — Max (seconds) between threads |
| Results / prompt | All | 1–4 files per prompt |
| Resolution | T2V / I2V only | 480p / 720p (images output 1K, auto-derived by ratio) |
| Reference Image Folder | I2I & I2V | Auto-matches images by keyword in the prompt |
| Save Folder | All | Output directory |
Text-to-Image Tab (T2I)
Generate a 1K image from a text prompt — no reference images needed.
- Enter prompts (one per line) or import from TXT/Excel
- Pick an aspect ratio + results per prompt
- Click 🚀 RUN
Text-to-Video Tab (T2V)
Generate videos from text prompts at 480p/720p.
- Enter prompts
- Pick resolution + aspect ratio
- Click 🚀 RUN
Image-to-Image Tab (I2I)
Generate images from up to 3 named components: Character/Subject, Scene, Style (at least one required).
- Drag-and-drop images into the Character / Scene / Style slots
- Enter a prompt describing the desired result
- Click 🚀 RUN
I2I supports auto-match: set a reference image folder in the config → the app maps images whose filename matches a keyword in the prompt onto the slots.
Image-to-Video Tab (I2V)
Generate video from frames (a start image + optional end image for interpolation) or from composed component images.
- Pick the reference mode: Frames (start/end) or Components
- Load images + enter a prompt describing the motion
- Pick resolution / aspect ratio
- Click 🚀 RUN
Action Buttons
| Button | Function |
|---|---|
| 🚀 RUN | Process every Pending row in the current tab |
| ▶️ Run Selected | Only run rows you have ticked |
| 🔄 Retry Errors | Re-run rows currently in Error / Failed state |
| ⏸️ PAUSE | Pause — running tasks finish, no new tasks dispatched |
| ⏹️ STOP | Stop all processing |
| 📂 Load Session | Restore prompts + results from a previous session |
The Meta AI page gates Run / Run Selected / Retry Errors on plan tier. Basic-tier accounts see the Plus/Max popup.