Bot Command Guide

How to Create Your First Video

1. Generate a base image: Use /generate to create your starting image
2. Reply with motion prompt: Reply to the image with /i2v [your motion description]
3. Wait for processing: You'll see live status updates as your video progresses through each stage
4. Receive your video: The final MP4 will be sent as a document you can download

💡 Pro Tips:

• Keep motion prompts focused on 1-2 main actions for best results
• Avoid requesting complex scene changes or teleportation
• The AI works best with natural, flowing movements
• Your daily limit resets at midnight (server time)

Step 1: Start the Bot & Choose a Nickname (/start)

When you first launch the bot, it greets you and asks for a nickname. This nickname cannot be changed later, so choose wisely.

• If you already have a nickname, the bot starts immediately and shows short info about the most important commands.

Step 2: Set a Default Prompt (/setprompt)

(OPTIONAL ONLY FOR Experienced Users)
You can define a custom prompt that will automatically be appended to each image generation. This way, you don't have to repeatedly add terms like high quality or masterpiece.

• Example: /setprompt masterpiece, high quality
  Now every prompt you enter via /generate will automatically include "masterpiece, high quality".
• To remove the stored prompt, type /setprompt 0.

Step 3: Exclude Undesired Elements (/setnegative)

A negative prompt excludes certain subjects or flaws from your images. This ensures that things like watermarks or poor quality won't appear.

Important! The bot has predefined negative prompts that already cover a wide range of undesirable elements, so additional negatives are often unnecessary.

• Example: /setnegative blurry, low quality, watermark
  The bot will do its best to avoid these three terms.

Step 4: Generate Your First Image (/generate / /gen)

Use /generate followed by your idea to get an image immediately. The bot uses default settings (e.g., 1024×1024 pixels) if you don't specify otherwise.

• Example: /generate A wolf on a mountain at sunset
  You'll get an image featuring a wolf in a mountainous sunset scene.
• Tip: use /gen to quickly repeat your last prompt — useful if you're not satisfied with the first result or want multiple variations.

Step 5: Review Your Settings (/settings)

Before diving into advanced configurations, it's helpful to check your current settings. The /settings command shows all your account parameters at once.

• Example: /settings displays your current model, sampler, steps, scale, size, seed, prompts, and all other configured values.
• This is especially useful to verify changes you've made or to share your setup with others for troubleshooting.

Step 6: Fine-Tune Your Settings (Size, Steps, Seed & More)

If you want more control over your images, you can use these various commands:

• /setsize 1024x1024 sets the image dimensions (width × height).
• /setsteps 30 increases the sampling steps. More steps = more details, but also longer wait times.
• /sampler lets you choose a specific algorithm (e.g., Euler a, DPM++). See below for sampler info.
• /setseed 12345 for reproducible results. -1 = random seed each time.
• /setscale 7.5 (CFG Scale) controls how strictly the bot follows your prompt. Too high can cause errors; too low may yield unwanted outcomes. A range of 4–8 is typical.

Step 7: Premium Features for More Control & Quality

Premium users gain access to additional commands for deeper refinement:

• /hires & /sethires for High-Resolution Fix and more detailed setup.
• /facefix automatically corrects faces.
• /lora applies specialized style or character LoRAs, letting you steer outputs (e.g., art styles, characters, furry looks, poses).
• /img2img transforms existing images using a new prompt.

Step 8: Upscaling Images (/up2 and /up4, Premium)

To increase the resolution of a generated image, reply to it with:

• /up2 doubles the resolution (2×).
• /up4 quadruples the resolution (4×), ideal for large formats.
• Note that upscaling uses credits, which are included monthly in Premium plans.

Step 9: Try the Restless Random Gambling Feature

Add an element of surprise to your prompts by using list placeholders:

• Add ?ListName? anywhere in your prompt (e.g., ?Dog?, ?Color?)
• The bot will randomly select an entry from that list and expand it with details
• Example: /gen ?Dog? in a fantasy garden might result in a Golden Retriever or a Siberian Husky with detailed characteristics
• Both Free and Premium users can use this feature, but Premium users get larger libraries with more detailed descriptions

Step 10: Help & Support

If something isn't working or you have questions, these commands can assist:

• /help displays the command overview again.
• /rules outlines all guidelines and restrictions.
• /getmyid shows your user ID, which an admin might need for support.
• /settings shows all your current configuration values for easy reference.

How to Use LoRAs with the Bot

Below is a quick guide on enabling/disabling LoRAs, defining triggers, and adjusting their weights via the bot's commands.

• /lora – Opens a LoRA selection menu, showing which LoRAs are available. You can enable or disable them here. Some LoRAs will automatically add a known trigger word to your prompt; others require manual addition.
• Enable a LoRA: Select it from the menu
  
• Disable a LoRA: Either deselect it in the /lora menu
• Adjust Weights: A LoRA's weight determines how strongly it affects your image. A value of 1.0 is considered neutral, anything above increases the LoRA's influence, and anything below reduces it.
  The GIF shows exactly how to specify and adjust these weights in your prompt.
• Enable a TriggerWord: Some LoRAs define one or more keywords that activate their unique styles or character details when included in your prompt. If the bot's menu shows these trigger words, simply select or deselect the preferred keywords.
  
  The GIF shows how to select or add the trigger word so the LoRA takes effect.
• Combine LoRAs: You can use multiple LoRAs in one generation as long as they don't conflict or override each other.
  The GIF shows how to select or add the trigger word so the LoRA takes effect.