POST /api/v1/{endpoint}     →  {"request_id": "abc123"}
             ↓
GET  /api/v1/predictions/{id}/result  →  poll until completed
             ↓
         {"status": "completed", "outputs": ["https://cdn.muapi.ai/..."]}

Generate Image

bash /mnt/skills/user/muapi-generate/scripts/generate-image.sh [options]

Models (`--model`)

Flag	Description
`flux-dev`	Flux Dev — fast, high quality (default)
`flux-schnell`	Flux Schnell — fastest
`flux-kontext-dev`	Flux Kontext Dev T2I
`flux-kontext-pro`	Flux Kontext Pro T2I
`flux-kontext-max`	Flux Kontext Max T2I
`hidream-fast`	HiDream i1 Fast
`hidream-dev`	HiDream i1 Dev
`hidream-full`	HiDream i1 Full
`midjourney`	Midjourney v7
`gpt4o`	GPT-4o image
`seedream`	ByteDance Seedream
`reve`	Reve
`qwen`	Qwen Image
`wan`	Wan 2.1 T2I

Examples

# Basic image (waits for completion)
bash generate-image.sh --prompt "A serene mountain landscape at sunrise"

# Midjourney v7
bash generate-image.sh --prompt "cinematic portrait, golden hour" --model midjourney

# GPT-4o with instruction-following
bash generate-image.sh --prompt "A red car with blue wheels on a white background" --model gpt4o

# Custom dimensions
bash generate-image.sh --prompt "a cat" --width 1280 --height 720

# Multiple images
bash generate-image.sh --prompt "abstract art" --num-images 4

# Async (returns request_id immediately)
bash generate-image.sh --prompt "a galaxy" --model hidream-full --async

Arguments

Argument	Description	Default
`--prompt`, `-p`	Text description (required)	—
`--model`, `-m`	Model flag (see table above)	`flux-dev`
`--width`	Image width in pixels	`1024`
`--height`	Image height in pixels	`1024`
`--aspect-ratio`	`1:1`, `16:9`, `9:16`, `4:3`, `3:4`	—
`--num-images`	Number of images (1–4)	`1`
`--async`	Submit and return request_id immediately	—
`--timeout`	Max seconds to wait	`300`
`--json`	Output raw JSON only	—

Generate Video (Text-to-Video)

bash /mnt/skills/user/muapi-generate/scripts/generate-video.sh [options]

Models (`--model`)

Flag	Description
`veo3`	Google Veo3 — highest quality
`veo3-fast`	Google Veo3 Fast
`kling-master`	Kling v2.1 Master T2V
`wan2`	Wan 2.1 T2V
`wan22`	Wan 2.2 T2V
`wan22-fast`	Wan 2.2 5B Fast T2V
`seedance-pro`	Seedance Pro T2V
`seedance-lite`	Seedance Lite T2V
`hunyuan`	HunyuanVideo T2V
`hunyuan-fast`	HunyuanVideo Fast T2V
`runway`	Runway T2V
`pixverse`	Pixverse v4.5 T2V
`vidu`	Vidu v2.0 T2V
`minimax-std`	Minimax Hailuo-02 Standard T2V
`minimax-pro`	Minimax Hailuo-02 Pro T2V (default)

Examples

# Default (waits for completion — may take 1–3 mins)
bash generate-video.sh --prompt "ocean waves crashing on a rocky shore"

# Veo3 — highest quality
bash generate-video.sh --prompt "a hummingbird in slow motion" --model veo3

# Kling — fast, reliable
bash generate-video.sh --prompt "a spaceship launching" --model kling-master

# Async mode (recommended for long videos)
bash generate-video.sh --prompt "epic battle scene" --model veo3 --async
# → Request ID: abc123-def456
# → Check: bash check-result.sh --id "abc123-def456"

# 9:16 vertical
bash generate-video.sh --prompt "city timelapse" --model wan2 --aspect-ratio 9:16

Queue Operations

# Submit and return immediately
bash generate-video.sh --prompt "..." --model veo3 --async

# Check status / get result
bash generate-video.sh --status "request_id" --model veo3
bash generate-video.sh --result "request_id"

Arguments

Argument	Description	Default
`--prompt`, `-p`	Text description (required)	—
`--model`, `-m`	Model flag	`minimax-pro`
`--aspect-ratio`	`16:9`, `9:16`, `1:1`	`16:9`
`--duration`	Duration in seconds (5 or 10)	`5`
`--async`	Submit and return request_id immediately	—
`--status ID`	Check status of a queued request	—
`--result ID`	Get result of a completed request	—
`--timeout`	Max seconds to wait	`600`
`--json`	Output raw JSON only	—

Image to Video

bash /mnt/skills/user/muapi-generate/scripts/image-to-video.sh [options]

Models (`--model`)

Flag	Description
`kling-std`	Kling v2.1 Standard I2V
`kling-pro`	Kling v2.1 Pro I2V (best overall)
`kling-master`	Kling v2.1 Master I2V
`veo3`	Google Veo3 I2V
`veo3-fast`	Google Veo3 Fast I2V
`wan2`	Wan 2.1 I2V
`wan22`	Wan 2.2 I2V
`seedance-pro`	Seedance Pro I2V
`seedance-lite`	Seedance Lite I2V
`hunyuan`	HunyuanVideo I2V
`runway`	Runway I2V
`pixverse`	Pixverse v4.5 I2V
`vidu`	Vidu v2.0 I2V
`midjourney`	Midjourney v7 I2V
`minimax-std`	Minimax Hailuo-02 Standard I2V
`minimax-pro`	Minimax Hailuo-02 Pro I2V

Examples

# From URL
bash image-to-video.sh \
  --image-url "https://example.com/photo.jpg" \
  --prompt "camera slowly zooms in" \
  --model kling-pro

# From local file (auto-uploaded to CDN)
bash image-to-video.sh \
  --file "/path/to/photo.jpg" \
  --prompt "gentle wind blowing through the scene" \
  --model veo3-fast

# With last frame (start+end interpolation)
bash image-to-video.sh \
  --image-url "https://example.com/start.jpg" \
  --last-image-url "https://example.com/end.jpg" \
  --model kling-pro

# Async
bash image-to-video.sh --image-url "..." --prompt "..." --model veo3 --async

Arguments

Argument	Description	Default
`--image-url`	Input image URL	—
`--file`, `--image`	Local file (auto-uploads to CDN)	—
`--last-image-url`	End frame URL (for start+end I2V)	—
`--prompt`, `-p`	Motion description	`""`
`--model`, `-m`	Model flag	`kling-pro`
`--aspect-ratio`	`16:9`, `9:16`, `1:1`	`16:9`
`--duration`	Duration in seconds (5 or 10)	`5`
`--async`	Submit and return request_id immediately	—
`--status ID`	Check status	—
`--result ID`	Get result	—
`--timeout`	Max seconds to wait	`600`

File Upload

bash /mnt/skills/user/muapi-generate/scripts/upload.sh --file "/path/to/file.jpg"

Returns the CDN URL for use in subsequent requests.

URL=$(bash upload.sh --file "/path/to/photo.jpg")
# → https://cdn.muapi.ai/files/abc123/photo.jpg

bash image-to-video.sh --image-url "$URL" --prompt "zoom in slowly"

Supported types: jpg, jpeg, png, gif, webp, mp4, mov, webm, mp3, wav

Output Format

Image Response

{
  "request_id": "abc123",
  "status": "completed",
  "outputs": ["https://cdn.muapi.ai/files/abc123/image.png"]
}

Video Response

{
  "request_id": "abc123",
  "status": "completed",
  "outputs": ["https://cdn.muapi.ai/files/abc123/video.mp4"]
}

Present Results to User

Images:

![Generated Image](https://cdn.muapi.ai/files/.../image.png)
• Model: Flux Dev | Generated in 3.2s

Videos:

[View Video](https://cdn.muapi.ai/files/.../video.mp4)
• Model: Kling Pro | Duration: 5s | Generated in 45s

Async Submission:

Request submitted to queue.
• Request ID: abc123-def456
• Model: veo3
• Check: bash check-result.sh --id "abc123-def456"

Troubleshooting

API Key Error

Error: MUAPI_KEY not set
Run: bash setup.sh --add-key "your_key_here"
Or:  export MUAPI_KEY=your_key_here

Timeout

Error: Timeout after 600s
Request ID: abc123-def456
Check manually: bash check-result.sh --id "abc123-def456"

Generate Image

Models (--model)

Examples

Arguments

Generate Video (Text-to-Video)

Models (--model)

Examples

Queue Operations

Arguments

Image to Video

Models (--model)

Examples

Arguments

File Upload

Output Format

Image Response

Video Response

Present Results to User

Troubleshooting

API Key Error

Timeout

Models (`--model`)

Models (`--model`)

Models (`--model`)

Script	Purpose
`generate-image.sh`	Text-to-image generation
`generate-video.sh`	Text-to-video generation (async)
`image-to-video.sh`	Image-to-video generation (async)
`upload.sh`	Upload local files to muapi CDN

muapi-generate

muapi.ai Generate

Scripts

API Flow