🧠 Overview

AI generation times can vary depending on the model, the complexity of your prompt, and server load.
In some cases, a generation may appear stuck, frozen, or take longer than expected, especially for video models like Kling, Veo, or WAN.

This article explains why this issue occurs, what to do when it happens, and how to improve speed and reliability.

---

⚙️ Why Generations May Be Slow

Cause	Description	Typical Fix
Model Queue (Server Load)	Popular models have limited simultaneous slots	The browser tab out of memory or lagging
Large Output Size	Higher resolution = longer rendering time	Lower resolution or duration
Complex Prompt	Too many objects, actions, or lighting details	Simplify your prompt
Network Instability	Slow or unstable connection interrupts progress	Refresh your network or switch Wi-Fi
Browser Resource Limit	Early-release models can run more slowly	Close other tabs and retry
Beta or Experimental Model	Early-release models can run slower	Use stable versions for speed

💡 Models like Veo 3, Kling 2.1, and WAN 2.5 may take 30–90 seconds longer for cinematic or long video scenes.

---

🔄 How the Generation Queue Works

When traffic is high, your request is added to a queue.
You’ll see a message like:

“Your generation is queued. Estimated start time: 30 seconds.”

The queue system ensures:

• Fair distribution of model access
• Stable server performance for all users
• Automatic retry if the model doesn’t respond in time

---

🚀 Improving Generation Speed

Here are ways to optimize speed and avoid stuck tasks:

🧠 Prompt Optimization

• Keep prompts under 800 characters.
• Avoid repeating unnecessary adjectives.
• Use clear, simple descriptions with key visual terms.

⚙️ Performance Adjustments

• Use lower resolution for drafts (e.g., 512×512 or 720p).
• Reduce duration for video tools (e.g., 4s–6s).
• Choose “Standard” quality instead of “High.”

🌐 Device & Browser Tips

• Use Google Chrome or Edge for the best performance.
• Close unused browser tabs or background apps.
• Avoid VPNs or unstable networks.

💡 For complex generations (e.g., cinematic videos or multi-scene prompts), it’s normal for render times to exceed 2–3 minutes.

---

❌ What to Do If It’s Stuck

If a generation stays frozen or doesn’t complete after several minutes:

1. If marked as “Failed,” click Retry Generation.
2. If still unresponsive after retrying, contact support.

⚠️ Do not reload the page mid-generation — this can cancel the process and delay refunds.

---

🧩 When to Contact Support

If your generation:

• Remains stuck for over 10 minutes
• Deducted credits but never completed
• Shows “Processing” indefinitely

Contact [email protected] with:

• Your Username
• Generation ID (from History tab)
• The model used (e.g., Kling 2.1, Veo 3, WAN 2.5)
• Date/time of issue

Include the subject: “Stuck Generation – [Your Username]”

---

⚙️ Troubleshooting Checklist

✅ Check model status banner (may be temporarily under maintenance)
✅ Lower output size or duration
✅ Retry once from History tab
✅ Clear browser cache or switch device
✅ Contact support if issue persists