Action Tags

Action tags are special commands in Treechat that let you invoke AI models and services directly from your posts. Simply type a command starting with ! followed by your prompt, and Treechat will generate content for you.

What are Action Tags?

Action tags are commands that start with ! (exclamation mark) that trigger AI processing. For example:

• !flux create a sunset over mountains - Generates an AI image
• !edit_nb make the sky more dramatic - Edits an image earlier in the thread
• !veo3 a bird flying through clouds - Generate a video
• !animate bring this image to life - Animates an image from earlier in the thread

How to Use Action Tags

1. Type the command: Start your post with ! followed by the action tag name
2. Add your prompt: Write what you want the AI to create or do
3. Post it: Submit your post like normal
4. Wait for results: The AI will process your request and attach the result to your post

Example

!flux a beautiful cyberpunk cityscape at night

This will generate an AI image of a cyberpunk city and attach it to your post.

Image Generation

Quality Tip: For the best quality image generation, use !nb (Nano Banana / Gemini 2.5 Flash) - it’s currently the highest quality image generator available.

!nb

Generate images with Google’s Gemini 2.5 Flash (aka “Nano Banana”) - best quality image generator available.

Usage: !nb [your image description]

Examples:

• !nb a serene mountain landscape at sunset
• !nb portrait of a futuristic robot
• !nb abstract geometric patterns in vibrant colors

Cost: ~$0.09 private, FREE public

Note: This is currently the highest quality image generation model in Treechat.

!flux

Generate images with Flux.1 by Black Forest Labs - fast and high quality.

Usage: !flux [your image description]

Examples:

• !flux a cat wearing a space suit
• !flux photorealistic portrait of an astronaut
• !flux abstract art with vibrant colors

Cost: ~$0.09 per image (private), FREE (public)

!qwen

Generate images with Qwen - advanced image generation with excellent text rendering capabilities.

Usage: !qwen [your image description]

Examples:

• !qwen a neon sign saying "Hello World"
• !qwen poster design with the text "AI Art"

Cost: ~$0.05 per image (private), FREE (public)

Note: Qwen excels at generating images with complex text and typography.

!dalle

Generate images with DALL-E 3 by OpenAI.

Usage: !dalle [your image description]

Example: !dalle a cozy coffee shop in autumn

Cost: ~$0.09 per image (private), FREE (public)

!gpt4o

Generate images using OpenAI’s GPT-4o multimodal image generation.

Usage: !gpt4o [your image description]

Example: !gpt4o realistic photo of a mountain sunrise

Cost: ~$0.30 per image

Video Generation

!veo3

Generate high-quality videos with Google Veo 3 - best quality video generator available.

Usage: !veo3 [your video description]

Examples:

• !veo3 a bird flying through clouds
• !veo3 cinematic shot of waves crashing on beach

Output: High-quality video Cost: ~$7.99 per video

Note: This is currently the highest quality video generation model available in Treechat.

!veo2

Generate high-quality videos with Google Veo 2 - can be used for both text-to-video and image-to-video.

Usage: !veo2 [your video description]

Examples:

• !veo2 a person walking down a city street
• !veo2 aerial view of mountains at sunrise

Output: 5-second video Cost: ~$2.99 per video

Note: Veo 2 can generate videos from text prompts and can also animate existing images (see !animate_veo below)

!luma

Generate photorealistic videos with Luma Dream Machine - can be used for both text-to-video and image-to-video.

Usage: !luma [your video description]

Example: !luma waves crashing on a beach at sunset

Output: 5-second video Cost: ~$0.99 per video

Note: Luma can generate videos from text prompts and can also animate existing images (see !animate_luma below)

!kling

Generate high-quality videos with Kling v2.1 - can be used for both text-to-video and image-to-video.

Usage: !kling [your video description]

Examples:

• !kling a beautiful sunset timelapse
• !kling dancing cat in a disco

Output: 5-second videos Cost: ~$1.99 per video

Note: Kling can generate videos from text prompts and can also animate existing images (see !animate_kling below)

!wan

Generate videos with Alibaba WAN 2.5 - can be used for both text-to-video and image-to-video.

Usage: !wan [your video description]

Example: !wan a serene mountain landscape with flowing water

Note: WAN can generate videos from text prompts and can also animate existing images (see !animate_wan below)

Output: 5 seconds at 480p Cost: ~$0.99 per video

!seedance

Generate videos with ByteDance SeedAnce-1-Pro - can be used for both text-to-video and image-to-video.

Usage: !seedance [your video description]

Example: !seedance a person walking through a misty forest

Output: 5 seconds at 720p, 24fps Cost: ~$0.99 per video

Note: SeedAnce can generate videos from text prompts and can also animate existing images (see !animate_seedance below)

!sora2

Generate high-quality videos with OpenAI Sora 2 - can be used for both text-to-video and image-to-video.

Usage: !sora2 [your video description]

Examples:

• !sora2 a drone flying over a mountain valley
• !sora2 time-lapse of clouds moving across the sky

Output: High-quality video Cost: ~$0.99 per video

Note: Sora 2 can generate videos from text prompts and can also animate existing images (see !animate_sora2 below)

!minimax

Generate videos with Minimax/Hailuo - can be used for both text-to-video and image-to-video.

Usage: !minimax [your video description]

Example: !minimax a butterfly landing on a flower

Output: 5-second video Cost: ~$0.99 per video

Note: Minimax can generate videos from text prompts and can also animate existing images (see !animate below)

Audio & Music

!suno

Generate complete songs with Suno AI.

Usage: !suno [song description]

Example: !suno create a jazz song about coffee and morning vibes

Remix Actions (Editing Existing Images)

Remix actions let you modify images that appear earlier in the conversation thread. Instead of generating new content, they transform existing images based on your prompt.

Quality Tip: For the best quality image editing, use !edit_nb (Nano Banana / Gemini 2.5 Flash) - it’s currently the highest quality image editor available.

!edit_nb

Edit images using Google’s Gemini 2.5 Flash (aka “Nano Banana”) - best quality image editor available.

Usage: !edit_nb [editing instructions]

Example: !edit_nb add snow to the landscape

Note: Requires an image earlier in the thread to edit.

Cost: ~$0.09 per edit

!edit

Edit images using Google’s Gemini Flash model.

Usage: !edit [editing instructions]

Example: !edit make the sky more dramatic

Note: Requires an image earlier in the thread to edit.

Cost: ~$0.09 per edit

!edit_openai

Edit images using OpenAI’s GPT-4o multimodal model.

Usage: !edit_openai [editing instructions]

Example: !edit_openai remove the person from this image

Note: Requires an image earlier in the thread to edit.

Cost: ~$0.30 per edit

!edit_flux

Edit images using Flux Kontext Pro - state-of-the-art image editing with excellent prompt following.

Usage: !edit_flux [editing instructions]

Example: !edit_flux change the lighting to golden hour

Note: Requires an image earlier in the thread to edit.

Cost: ~$0.09 per edit

!edit_qwen

Edit images using Qwen with img2img pipeline.

Usage: !edit_qwen [editing instructions]

Example: !edit_qwen enhance the colors

Note: Requires an image earlier in the thread to edit.

Cost: ~$0.05 per edit

!animate (Minimax)

Generate videos using an earlier image as the first frame with Minimax/Hailuo.

Usage: !animate [animation description]

Example: !animate make this image come to life with gentle movement

Note: Requires an image earlier in the thread as the starting frame.

Output: 5-second video Cost: ~$0.99 per video

!animate_kling

Generate videos using an earlier image as the first frame with Kling v2.1.

Usage: !animate_kling [animation description]

Example: !animate_kling add realistic motion to this scene

Note: Requires an image earlier in the thread as the starting frame.

Output: 5-second video Cost: ~$1.99 per video

!animate_luma

Generate videos using an earlier image as the first frame with Luma Dream Machine.

Usage: !animate_luma [animation description]

Example: !animate_luma animate this with smooth camera movement

Note: Requires an image earlier in the thread as the starting frame.

Output: 5-second video Cost: ~$0.99 per video

!animate_veo

Generate videos using an earlier image as the first frame with Google Veo 2.

Usage: !animate_veo [animation description]

Example: !animate_veo bring this scene to life

Note: Requires an image earlier in the thread as the starting frame.

Output: 5-second video Cost: ~$2.99 per video

!animate_veo3

Generate videos using an earlier image as the first frame with Google Veo 3 - best quality image-to-video available.

Usage: !animate_veo3 [animation description]

Example: !animate_veo3 bring this scene to life with cinematic motion

Note: Requires an image earlier in the thread as the starting frame.

Output: High-quality video Cost: ~$7.99 per video

!animate_seedance

Generate videos using an earlier image as the first frame with ByteDance SeedAnce-1-Pro.

Usage: !animate_seedance [animation description]

Example: !animate_seedance add dynamic movement

Note: Requires an image earlier in the thread as the starting frame.

Output: 5 seconds at 720p, 24fps Cost: ~$0.99 per video

!animate_sora2

Generate videos using an earlier image as the first frame with OpenAI Sora 2.

Usage: !animate_sora2 [animation description]

Example: !animate_sora2 animate this with cinematic camera movement

Note: Requires an image earlier in the thread as the starting frame.

Output: High-quality video Cost: ~$0.99 per video

!animate_wan

Generate videos using an earlier image as the first frame with Alibaba WAN 2.5.

Usage: !animate_wan [animation description]

Example: !animate_wan add gentle motion to this scene

Note: Requires an image earlier in the thread as the starting frame.

Output: 5 seconds at 480p Cost: ~$0.99 per video

AI Assistants

In addition to action tags for generating media, Treechat also supports conversational AI assistants using @@ mentions. These let you chat with language models like Claude Sonnet 4.5 (default) and GPT-4o for questions, analysis, and discussions.

Learn more about AI Assistants →

Costs and Billing

Action tags that use AI models have costs associated with them:

• Costs are deducted from your BSV wallet
• Prices vary by model (typically $0.03 - $1.40 per generation)
• Public posts may have different pricing than private posts
• Check your wallet balance before using expensive models

Make sure you have funded your wallet - See Funding Your Wallet for details.

Tips for Best Results

For Image Generation

• Be specific: Describe colors, style, mood, and composition
• Use style keywords: “photorealistic”, “watercolor”, “cyberpunk”, “minimalist”
• Mention lighting: “golden hour”, “dramatic lighting”, “soft shadows”
• Try different models: Each AI has different strengths

For Video Generation

• Keep it simple: Complex scenes may not generate well
• Describe motion: “slow zoom”, “camera panning”, “rotating”
• Set the scene: Describe environment and atmosphere
• Be patient: Video generation can take several minutes

For Image Editing (Remix Actions)

• Be specific: Describe exactly what you want to change
• Reference the original: Mention specific elements from the source image
• Use clear instructions: “add”, “remove”, “change”, “enhance”
• Try different models: Each has different editing strengths

For Image-to-Video (Animate Actions)

• Start with good images: Better source images produce better videos
• Describe the motion: Be specific about how things should move
• Keep it realistic: Natural motions work better than complex changes
• Consider the model: Different models excel at different types of animation

For AI Conversations (@@ Mentions)

• Ask clear questions: Be specific about what you need
• Provide context: Give background information when needed
• Continue the conversation: Follow up with additional questions
• Reference chat history: The AI remembers your previous messages

Combining Action Tags

You can chain action tags together in conversation threads to create complex workflows:

Example workflow:

1. Generate an image: !nb a beautiful mountain landscape
2. Edit the result: !edit_nb add a sunset in the background
3. Animate it: !animate_veo3 gentle camera pan across the scene
4. Discuss it: Reply with @@sonnet what are some ways to improve landscape photography?

This allows you to iteratively refine and transform content through multiple AI models.

Note: Currently, AI assistants invoked with @@ mentions cannot see images or videos in the thread - they only read text. Use them for questions, analysis, and discussions based on text content.

Troubleshooting

Action tag didn’t work?

• Check that you started with ! (exclamation mark)
• Make sure you spelled the tag name correctly
• Verify you have sufficient BSV balance for paid models

Generation failed?

• Some prompts may be filtered for content policy
• Check if the model is temporarily unavailable
• Try rephrasing your prompt

No result appearing?

• Large videos/images may take several minutes to generate
• Check your notifications for completion status
• Refresh the page to see the latest results

Next Steps

• Funding Your Wallet - Add BSV to use action tags
• Understanding Upvalue - Earn BSV from your AI-generated content
• Searching Treechat - Find others using action tags