Skip to main content

​
nut resource

Manage and generate resources (images, audio, video) using AI. 
nut resource [command]
Aliases: resource, resources

​
Overview

The nut resource command enables you to manage and generate media assets. Upload files, manage metadata, and generate images, audio, and video using AI models directly from the command line.

​
nut resource list

List all resources in your Coconut instance. 
nut resource list [options]
nut resource ls [options]
Aliases: list, ls Options:
  • -s, --search <query> - Filter resources by name, description, or tags
  • --type <type> - Filter by media type prefix (e.g. image, audio, video)
  • --json - Output as JSON
  • --api <url> - API server URL (default: “http://localhost”)
Examples: 
# List all resources
nut resource list

# Search for resources
nut resource ls --search "logo"

# Filter by type
nut resource list --type image --json

​
nut resource get

Get details for a specific resource. 
nut resource get <id> [options]
nut resource show <id> [options]
Aliases: get, show Arguments:
  • id - Resource ID
Options:
  • --json - Output as JSON
  • -o, --output <path> - Download resource file to local path
  • --api <url> - API server URL (default: “http://localhost”)
Examples: 
# Get resource details
nut resource get res-1234567890-abc123def

# Get resource as JSON
nut resource show res-1234567890-abc123def --json

# Download resource file
nut resource get res-1234567890-abc123def -o ./downloaded-file.png

​
nut resource add

Upload a file to resources. 
nut resource add <file...> [options]
nut resource upload <file...> [options]
Aliases: add, upload Arguments:
  • file - File path(s) to upload
Options:
  • -t, --tags <tags> - Comma-separated tags
  • -d, --description <text> - Description of the resource
  • --json - Output as JSON
  • --api <url> - API server URL (default: “http://localhost”)
Examples: 
# Upload a single file
nut resource add ./screenshot.png

# Upload with tags and description
nut resource upload ./logo.svg -t "branding, logo" -d "Company logo"

# Upload multiple files
nut resource add ./image1.png ./image2.png -t "batch"

# Upload and output as JSON
nut resource add ./data.json --json

​
nut resource update

Update resource metadata (tags, description). 
nut resource update <id> [options]
Arguments:
  • id - Resource ID to update
Options:
  • -t, --tags <tags> - Comma-separated tags (replaces existing)
  • -d, --description <text> - Description (replaces existing)
  • --json - Output as JSON
  • --api <url> - API server URL (default: “http://localhost”)
Examples: 
# Update tags
nut resource update res-123 -t "logo, branding"

# Update description
nut resource update res-123 -d "Updated description"

# Update both tags and description
nut resource update res-123 -t "new-tag" -d "New desc" --json

​
nut resource delete

Delete a resource. 
nut resource delete <id> [options]
nut resource rm <id> [options]
Aliases: delete, rm Arguments:
  • id - Resource ID to delete
Options:
  • -f, --force - Skip confirmation prompt
  • --json - Output as JSON
  • --api <url> - API server URL (default: “http://localhost”)
Examples: 
# Delete a resource
nut resource delete res-1234567890-abc123def

# Delete without confirmation
nut resource rm res-1234567890-abc123def --force

# Delete and output as JSON
nut resource delete res-123 --json
Deleting a resource is permanent. Make sure you want to remove it before confirming.

​
nut resource generate-image

Generate an image using AI models via Replicate. 
nut resource generate-image <prompt> [options]
nut resource image <prompt> [options]
Aliases: generate-image, image Arguments:
  • prompt - Description of the image to generate
Options:
  • -m, --model <model> - Model to use (default: nano-banana-pro)
  • --save - Save generated image to resources (.nut/resources/)
  • -o, --output <path> - Download image to local file path
  • --api <url> - API server URL (default: “http://localhost”)
Models:
CLI flagReplicate modelNotes
nano-banana-progoogle/nano-banana-proDefault. Supports resolution (1K/2K/4K), image input for style reference
flux-2-devblack-forest-labs/flux-2-devSupports --go-fast, quality (60-100), image input
flux-2-problack-forest-labs/flux-2-proSupports resolution (0.5 MP–4 MP), --safety (1-6), image input
gpt-image-1.5openai/gpt-image-1.5Supports --quality (low/medium/high), --background (auto/transparent/opaque), limited aspect ratios (1:1, 3:2, 2:3)
Examples: 
# Default model (nano-banana-pro)
nut resource image "A serene morning in Tokyo"

# Generate and save to resources
nut resource image "A serene morning in Tokyo" --save

# Use flux-2-pro and save to resources
nut resource image "A serene morning in Tokyo" -m flux-2-pro --save

# Use gpt-image-1.5 with high quality and save to file
nut resource image "Abstract art" -m gpt-image-1.5 --quality high -o ./art.png

# Generate and download to specific location
nut resource image "Product mockup for mobile app" --output ./assets/mockup.png
Be specific in your prompts. Include details about style, composition, lighting, and mood for better results.

​
nut resource generate-audio

Generate audio from text using AI text-to-speech. 
nut resource generate-audio <text> [options]
nut resource audio <text> [options]
Aliases: generate-audio, audio, tts Arguments:
  • text - Text to convert to speech
Options:
  • -v, --voice <id> - Voice ID (default: “English_Trustworth_Man”)
  • -s, --speed <number> - Speech speed multiplier, 0.5-2.0 (default: “1”)
  • -p, --pitch <number> - Pitch adjustment, -12 to 12 (default: “0”)
  • -e, --emotion <type> - Emotion style (auto, happy, sad, angry, fearful, surprised, disgusted) (default: “auto”)
  • -l, --language <lang> - Language boost (default: “English”)
  • -f, --format <format> - Audio format (mp3, wav, ogg) (default: “mp3”)
  • --save - Save generated audio to resources (.nut/resources/)
  • -o, --output <path> - Download audio to local file path
  • --api <url> - API server URL (default: “http://localhost”)
Model: Uses minimax/speech-02-turbo for high-quality text-to-speech generation. Examples: 
# Generate audio narration
nut resource audio "Hello, welcome to Coconut!"

# Generate with specific voice and speed
nut resource tts "Welcome!" -v English_Calm_Woman -s 1.2

# Generate and save to resources
nut resource audio "This is a test" --save -f wav
Generated audio is suitable for narration, voiceovers, podcast intros, and other audio content needs.

​
nut resource generate-video

Generate a video using AI. 
nut resource generate-video <prompt> [options]
nut resource video <prompt> [options]
Aliases: generate-video, video, vid Arguments:
  • prompt - Description of the video to generate
Options:
  • -d, --duration <seconds> - Video duration in seconds, 5-16 (default: “8”)
  • -r, --resolution <res> - Video resolution (480p, 720p, 1080p) (default: “720p”)
  • -a, --aspect-ratio <ratio> - Aspect ratio (16:9, 9:16, 1:1) (default: “16:9”)
  • --start-frame <id> - Resource ID of image to use as first frame
  • --end-frame <id> - Resource ID of image to use as last frame
  • --no-audio - Disable AI audio generation for the video
  • --save - Save generated video to resources (.nut/resources/)
  • -o, --output <path> - Download video to local file path
  • --api <url> - API server URL (default: “http://localhost”)
Model: Uses google/veo-3.1-fast for fast, high-quality video generation. Examples: 
# Generate a short video
nut resource video "A rocket launching into space"

# Generate with specific parameters
nut resource video "Ocean waves" -d 12 -r 1080p --save

# Generate with start and end frames
nut resource vid "Scene transition" --start-frame res-abc --end-frame res-xyz

# Generate silent video and save locally
nut resource video "Silent timelapse" --no-audio -o ./timelapse.mp4
Video generation takes longer than image or audio. Be patient and ensure you have a stable internet connection.

​
Use Cases

​
Documentation Assets

# Create diagrams and illustrations
nut resource image "System architecture diagram showing microservices" --save

# Upload existing screenshots
nut resource add ./screenshot.png -t "docs, tutorial" -d "Dashboard screenshot"

​
Marketing Materials

# Social media content
nut resource image "Eye-catching banner for product launch" --save

# Video ads
nut resource video "Product showcase with dynamic camera movements" -d 15 --save

​
Audio Content

# Podcast intros
nut resource audio "Welcome to the Tech Talk podcast, where we discuss the latest innovations" --save

# Tutorial narration
nut resource audio "Step one: Open the application and navigate to settings" -v English_Calm_Woman -s 1.1

​
Best Practices

The quality of generated resources depends heavily on prompt clarity. Include specific details about what you want, including style, mood, colors, composition, and any key elements.
If the first result isn’t perfect, refine your prompt and try again. Small changes in wording can produce significantly different results.
Use the --save option to store resources in your Coconut instance, or --output to save files locally in organized directories.
AI resource generation can be computationally expensive. Be mindful of the number and complexity of assets you generate, especially videos.

​
Model Information

  • Images: Generated via Replicate’s AI models
  • Audio: Uses MiniMax speech-02-turbo for natural-sounding speech
  • Video: Powered by Google’s Veo 3.1 Fast model for efficient video generation
Resource generation requires the Coconut server to be running (nut serve) and a Replicate API token configured via nut config set replicate <token>.