AI Content Generation
Postuby generates visuals, videos, headlines, and hashtags in your brand voice — one at a time or in bulk.
- Static visual — Feed post (1:1, 4:5), Story (9:16), banner (16:9)
- Carousel — Multi-slide post (2-10 slides). With "Carousel Plan" you can split a topic into N slides
- Reels / TikTok video — Script + scenes + optional AI voiceover
- Caption (headline & body) — On-brand post copy
- Hashtags — Industry- and language-specific trending sets
- Story — Snap-format visuals with sticker recommendations
For visuals: OpenAI gpt-image, Google Imagen, Kling AI, and Kie.AI (nano-banana). Which one runs depends on your plan, the chosen quality tier, and the visual type — the system picks automatically.
For text/captions/hashtags: OpenAI GPT and Google Gemini.
For video: VideoLab (scene-based composer) plus optional Veo or Kling video models.
Your brand's voice, colors, logo, industry, and target audience are auto-injected into every prompt. Even if you don't ask, fake logos are never generated (we have a dedicated guardrail).
For better results:
- Fill the brand description in detail (voice, visual references)
- Upload templates (in 1:1, 4:5, 9:16, 16:9 formats)
- Upload Style Match photos (reference images)
- Keep "Strict style" enabled when creating a plan
For a single post we generate 6 variations in one click — different compositions, color emphasis, and layouts. Pick the one you like; regenerate if needed.
Not enough? The "Regenerate" button on the post detail produces 6 more. Mark variations you don't like as "Negative feedback" and AI will steer away from similar styles in future generations.
A special plan type that breaks one topic into N slides. Example: ask for 5 slides on "Social media mistakes" and AI delivers:
- Slide 1: Cover (hook + question)
- Slides 2-4: 3 mistakes + how to fix each
- Slide 5: CTA (save / comment)
You can run K topics × N slides bulk generation in one go (e.g. 5 topics × 4 slides = 20 slides at once).
Two paths:
1. Scene-based composer (VideoLab): describe the topic or script, AI assembles scenes (visual + text overlay + transitions), optionally adds voiceover. Output is 9:16 vertical, 15-60 seconds.
2. Avatar Studio (beta): An AI avatar (face + voice) presents the script. Ideal for product intros, education, short explainers. Camera framing (close-up/medium/wide) is selectable.
Generation takes 30 seconds to 3 minutes depending on the model. You're notified when it's ready and can drop it into the calendar to publish.
- Sharpen brand voice — write a more detailed brand description, change voice setting
- Upload Style Match photos — add 3-5 reference images so AI learns the visual style
- Try a different aspect ratio — 1:1 vs 4:5 give different compositions
- Give negative feedback — mark variations you don't like (AI avoids similar)
- Regenerate — try another set of 6 variations
- Ask the assistant — natural-language tweaks like "make this more energetic", "keep colors vivid" work