AI Generation Issues

​

My AI generation failed — where are my credits?

Credits are automatically refunded when a generation fails. You don’t need to request a refund — it happens instantly. Check your credit balance by clicking your profile picture (top right) → Credits. The refunded amount matches exactly what was deducted for that generation.

​

Common failure reasons

​

Content blocked

Cause: The AI model’s safety filter detected potentially inappropriate content in your prompt. Fix:

• Rephrase your prompt to avoid sensitive topics
• Avoid requests for realistic depictions of real people
• Remove explicit or violent descriptions
• Try a different AI model — each model has different content policies

​

Generation timed out

Cause: The AI model took too long to generate. This is more common with video models and high-resolution images. Fix:

• Try again — timeouts are usually temporary server load issues
• For videos: try a shorter duration (e.g., 5 seconds instead of 10)
• For images: try a lower resolution or different aspect ratio
• Try during off-peak hours

​

Image reference issues

Cause: The reference image you provided couldn’t be processed — it may be too small, too large, or in an unsupported format. Fix:

• Use images between 512×512 and 4096×4096 pixels
• Use JPG, PNG, or WebP format
• Ensure the image file is not corrupted
• Try without the reference image to isolate the issue

​

No output generated

Cause: The AI model couldn’t produce a result from your prompt. This can happen with very abstract or contradictory prompts. Fix:

• Be more specific in your description
• Use the Prompt Optimizer (✨ button) to improve your prompt
• Try a different model that may interpret your prompt better

​

My image quality is poor

Tips for better results:

1. Use the Prompt Optimizer — Click the ✨ button to let AI enhance your prompt. This dramatically improves results.
2. Be specific — “A golden retriever puppy on a beach at sunset, professional photography, shallow depth of field” works better than “dog on beach.”
3. Choose the right model:
   • Photorealism: Nano Banana Pro, FLUX.2 Pro, GPT Image 2 (HD quality)
   • Illustrations/Art: Recraft V4, Ideogram V3
   • Fast drafts: GPT Image Mini, Nano Banana
   • Text in images: Ideogram V3 (best at rendering text accurately)
4. Aspect ratio matters — Use the aspect ratio that matches your target platform (9:16 for Stories/Reels, 1:1 for feed posts, 16:9 for YouTube).

​

My video has visual artifacts

Common causes:

• Very fast motion in the prompt — AI video models struggle with rapid movement
• Very detailed scenes with many small elements
• Unrealistic physics requests

Fix:

• Simplify the scene description
• Use prompts that describe slow, smooth motion
• Try a different video model (each has different strengths)
• For best quality, use Kling 2.0 Master or Veo 3.1

​

The AI generated wrong text in the image

Cause: Most AI models struggle with text rendering. They may misspell words or produce garbled text. Best options for text in images:

1. Use Ideogram V3 — it’s specifically designed for accurate text rendering
2. Use Recraft V4 — excellent for logos and design with text
3. Generate the image without text, then add text using the Photo Editor

​

I ran out of credits

Options to get more credits:

1. Purchase credit packages — Click your profile picture (top right) → Credits → choose a package
2. Wait for monthly reset — Paid plans receive fresh credits each billing cycle
3. Enable Auto-Recharge — Click your profile picture → Credits → Auto-Recharge tab, set a threshold to automatically purchase more credits when your balance gets low

​

Which model should I use?

​

Still having issues?

Contact support through the chat widget with:

• The model you used
• Your prompt (or a description of what you were trying to generate)
• The error message from the notification