AI Composer¶
The AI Composer turns a short description into an on-brand post — caption and a freshly generated image — for every platform you select. You write a prompt, pick your platforms, and press Generate; the work runs in the background so you can queue more while it finishes.
Write your prompt¶
Describe the post you want in the prompt box. A prompt can be up to 10,000 characters, and you can dictate it instead of typing using the microphone (voice dictation).
- Improve rewrites your prompt with AI, using your active brand's tone, palette, and verified details as context where they fit. If you don't like the result, Revert restores what you had.
- Buzzator saves your work as a draft automatically as you go.
Choose your platforms¶
Select one or more platforms for the post. The AI generates a caption and image tailored to each one. You can't press Generate until you have both a non-empty prompt and at least one platform selected — otherwise Buzzator tells you what's missing.
A few platform rules apply in AI mode:
| Platform | In the AI Composer |
|---|---|
| TikTok | You must pick a Privacy level (who can view the post) before generating. |
| YouTube | Video-only — it's rejected in AI mode. Use Video Studio or the Manual Composer. |
| Dribbble, Twitch | Manual Composer only — the AI Composer doesn't offer them. |
See per-platform post settings for format, aspect ratio, and carousel options.
Attach references and documents¶
Use the + menu to attach up to 10 items total from your device, your Media Library, or Canva — reference images and PDF documents share this single limit:
- Reference images — style and subject references. The AI studies their look and subject and generates a brand-new image; it never posts your reference as-is.
- Documents — PDFs are read as source material the post is written from (up to 3 are read), not as visuals.
Pick an image style¶
The Style control sets the look of the generated image for this post:
- Brand Kit (default) — your brand's own colors drive the image.
- A visual preset (for example, Bauhaus or Glacier) — overrides only the image's colors, mood, lighting, and finish.
Either way, your caption, brand voice, and compliance rules always follow the brand. A preset restyles the picture only. See Your Brand Kit for what "on brand" draws from.
Toggles¶
| Toggle | Effect |
|---|---|
| Include Logo | Places your brand logo in the generated image. |
| Made with AI | Overlays a "Made with AI" disclosure on the image. |
| Same image for all platforms | Generates one image and reuses it everywhere, auto-cropped to each platform's aspect ratio. Turn it off for a separate image per platform. TikTok always gets its own 9:16 version either way. |
Note
Same image for all platforms only appears — and only takes effect — when you have two or more platforms selected.
Generate and watch the queue¶
Press Generate and the job runs in the background. You don't have to wait on the page — closing the tab doesn't stop it, and the finished drafts land in your Posts Hub.
Progress shows on live queue cards:
- Done cards clear on their own after about 5 seconds.
- Failed cards stay put so you can Retry or Dismiss them — a failure never silently disappears.
- Stop aborts a generation that's still running.
Refine the result¶
After a post is generated you can adjust it before publishing:
- Refine the image with a follow-up instruction (for example, "make it brighter, add warm tones, remove background").
- Rewrite the caption with a follow-up prompt.
- Set alt text for each image.
Limits¶
Generation is metered so no single account or team monopolizes capacity:
| Limit | Value |
|---|---|
| Per user | 10 generations per minute |
| Per organization | 30 generations per minute |
| Monthly | A per-plan cap on AI posts |
If you hit the monthly cap, Buzzator returns a clear "limit reached" message telling you the current count and to upgrade. See Plans, limits & billing for your plan's allowance.