@felores/kie-ai-mcp-server 1.5.0 → 1.7.2

package/README.md

@@ -22,6 +22,7 @@ Access the world's best AI models through a single, developer-friendly API. Gene
 - **Nano Banana**: Lightning-fast image generation and editing
 - **ElevenLabs**: Studio-quality text-to-speech and sound effects
 - **ByteDance Seedance**: High-quality video with text-to-video and image-to-video
+- **ByteDance Seedream V4**: Advanced image generation and editing with unified interface
 
 ### 💰 **Affordable Pricing**
 Pay-as-you-go credit system means you only pay for what you use. Good for startups and enterprises looking to reduce AI costs.
@@ -340,24 +341,25 @@ Using explicit model (overrides default V5):
 **Note**: In custom mode, `style` and `title` are required. If `instrumental` is false, `prompt` is used as exact lyrics. The `callBackUrl` is optional and will use the `KIE_AI_CALLBACK_URL` environment variable if not provided. The `model` parameter defaults to "V5" but can be explicitly set to any available version.
 
 ### 9. `elevenlabs_tts`
-Generate speech from text using ElevenLabs multilingual TTS v2 model.
+Generate speech from text using ElevenLabs TTS models (Turbo 2.5 by default, with optional Multilingual v2 support).
 
 **Parameters:**
 - `text` (string, required): The text to convert to speech (max 5000 characters)
+- `model` (enum, optional): TTS model to use - "turbo" (faster, default) or "multilingual" (supports context)
 - `voice` (enum, optional): Voice to use - "Rachel", "Aria", "Roger", "Sarah", "Laura", "Charlie", "George", "Callum", "River", "Liam", "Charlotte", "Alice", "Matilda", "Will", "Jessica", "Eric", "Chris", "Brian", "Daniel", "Lily", "Bill" (default: "Rachel")
 - `stability` (number, optional): Voice stability (0-1, step 0.01, default: 0.5)
 - `similarity_boost` (number, optional): Similarity boost (0-1, step 0.01, default: 0.75)
 - `style` (number, optional): Style exaggeration (0-1, step 0.01, default: 0)
 - `speed` (number, optional): Speech speed (0.7-1.2, step 0.01, default: 1.0)
 - `timestamps` (boolean, optional): Whether to return timestamps for each word (default: false)
-- `previous_text` (string, optional): Text that came before current request for continuity (max 5000 chars)
-- `next_text` (string, optional): Text that comes after current request for continuity (max 5000 chars)
-- `language_code` (string, optional): ISO 639-1 language code for language enforcement (max 500 chars)
+- `previous_text` (string, optional): Text that came before current request (multilingual model only, max 5000 chars)
+- `next_text` (string, optional): Text that comes after current request (multilingual model only, max 5000 chars)
+- `language_code` (string, optional): ISO 639-1 language code for language enforcement (turbo model only, max 500 chars)
 - `callBackUrl` (string, optional): URL to receive task completion updates (uses KIE_AI_CALLBACK_URL environment variable if not provided)
 
 **Examples:**
 
-Basic TTS generation:
+Basic TTS generation (uses Turbo model by default):
 ```json
 {
   "text": "Hello, this is a test of the ElevenLabs text-to-speech system.",
@@ -365,86 +367,36 @@ Basic TTS generation:
 }
 ```
 
-Advanced voice controls:
-```json
-{
-  "text": "Welcome to our presentation on artificial intelligence",
-  "voice": "Aria",
-  "stability": 0.8,
-  "similarity_boost": 0.9,
-  "style": 0.3,
-  "speed": 1.1
-}
-```
-
-With continuity for longer texts:
-```json
-{
-  "text": "This is the second part of our conversation.",
-  "voice": "Roger",
-  "previous_text": "This is the first part of our conversation.",
-  "next_text": "This is the third part of our conversation."
-}
-```
-
-**Note**: The `callBackUrl` is optional and will use the `KIE_AI_CALLBACK_URL` environment variable if not provided. Generation typically takes 30 seconds to 2 minutes depending on text length.
-
-### 10. `elevenlabs_tts_turbo`
-Generate speech from text using ElevenLabs Turbo 2.5 TTS model (faster generation with language enforcement support).
-
-**Parameters:**
-- `text` (string, required): The text to convert to speech (max 5000 characters)
-- `voice` (enum, optional): Voice to use - "Rachel", "Aria", "Roger", "Sarah", "Laura", "Charlie", "George", "Callum", "River", "Liam", "Charlotte", "Alice", "Matilda", "Will", "Jessica", "Eric", "Chris", "Brian", "Daniel", "Lily", "Bill" (default: "Rachel")
-- `stability` (number, optional): Voice stability (0-1, step 0.01, default: 0.5)
-- `similarity_boost` (number, optional): Similarity boost (0-1, step 0.01, default: 0.75)
-- `style` (number, optional): Style exaggeration (0-1, step 0.01, default: 0)
-- `speed` (number, optional): Speech speed (0.7-1.2, step 0.01, default: 1.0)
-- `timestamps` (boolean, optional): Whether to return timestamps for each word (default: false)
-- `previous_text` (string, optional): Text that came before current request for continuity (max 5000 chars)
-- `next_text` (string, optional): Text that comes after current request for continuity (max 5000 chars)
-- `language_code` (string, optional): ISO 639-1 language code for language enforcement - Turbo 2.5 supports this feature (max 500 chars)
-- `callBackUrl` (string, optional): URL to receive task completion updates (uses KIE_AI_CALLBACK_URL environment variable if not provided)
-
-**Examples:**
-
-Fast TTS generation:
-```json
-{
-  "text": "This is a fast generation using the Turbo model.",
-  "voice": "Aria"
-}
-```
-
-With language enforcement:
+Fast generation with language enforcement (Turbo model):
 ```json
 {
   "text": "Bonjour, comment allez-vous?",
   "voice": "Rachel",
+  "model": "turbo",
   "language_code": "fr"
 }
 ```
 
-Advanced controls with continuity:
+Advanced voice controls with context (Multilingual model):
 ```json
 {
-  "text": "This is part two of our series.",
+  "text": "This is the second part of our conversation.",
   "voice": "Roger",
-  "stability": 0.9,
-  "similarity_boost": 0.8,
-  "previous_text": "This is part one of our series.",
-  "language_code": "en"
+  "model": "multilingual",
+  "stability": 0.8,
+  "similarity_boost": 0.9,
+  "previous_text": "This is the first part of our conversation.",
+  "next_text": "This is the third part of our conversation."
 }
 ```
 
-**Key Differences from Multilingual TTS:**
-- **Faster Generation**: Turbo 2.5 processes text 15-60 seconds (vs 30-120 seconds for multilingual)
-- **Language Enforcement**: Supports ISO 639-1 language codes for consistent language output
-- **Same Voice Options**: All 21 voices available
-- **Same Quality**: Maintains high audio quality with faster processing
+**Model Comparison:**
+- **Turbo 2.5** (default): Faster generation (15-60 seconds), supports language enforcement with `language_code`
+- **Multilingual v2**: Supports context with `previous_text`/`next_text`, generation takes 30-120 seconds
 
-**Note**: The `callBackUrl` is optional and will use the `KIE_AI_CALLBACK_URL` environment variable if not provided. Turbo 2.5 generation is faster and supports language enforcement.
+**Note**: The `callBackUrl` is optional and will use the `KIE_AI_CALLBACK_URL` environment variable if not provided. Choose Turbo model for speed and language enforcement, or Multilingual model for context-aware speech generation.
 
-### 11. `elevenlabs_ttsfx`
+### 10. `elevenlabs_ttsfx`
 Generate sound effects from text descriptions using ElevenLabs Sound Effects v2 model.
 
 **Parameters:**
@@ -564,6 +516,181 @@ Video with specific ending frame:
 
 **Note**: The `callBackUrl` is optional and will use the `KIE_AI_CALLBACK_URL` environment variable if not provided. Video generation typically takes 2-5 minutes depending on quality and complexity.
 
+### 13. `bytedance_seedream_image`
+Generate and edit images using ByteDance Seedream V4 models (unified tool for both text-to-image and image editing).
+
+**Parameters:**
+- `prompt` (string, required): Text prompt for image generation or editing (max 10000 chars)
+- `image_urls` (array, optional): Array of image URLs for editing mode (1-10 images, if not provided, uses text-to-image)
+- `image_size` (string, optional): Image aspect ratio (default: "1:1")
+  - Options: `1:1`, `4:3`, `3:4`, `16:9`, `9:16`, `21:9`, `9:21`, `3:2`, `2:3`
+- `image_resolution` (string, optional): Image resolution (default: "1K")
+  - `1K`: Standard resolution (1024px on shortest side)
+  - `2K`: High resolution (2048px on shortest side)
+  - `4K`: Ultra high resolution (4096px on shortest side)
+- `max_images` (integer, optional): Number of images to generate (1-6, default: 1)
+- `seed` (integer, optional): Random seed for reproducible results (default: -1 for random)
+- `callBackUrl` (string, optional): URL for task completion notifications
+
+**Examples:**
+
+Text-to-image generation:
+```json
+{
+  "prompt": "A majestic dragon perched atop a crystal mountain at sunset, digital art style",
+  "image_size": "16:9",
+  "image_resolution": "2K",
+  "max_images": 2,
+  "seed": 42
+}
+```
+
+Image editing:
+```json
+{
+  "prompt": "Transform the day scene into a magical night with glowing stars and moonlight",
+  "image_urls": ["https://example.com/day-landscape.jpg"],
+  "image_size": "16:9",
+  "image_resolution": "2K",
+  "max_images": 1
+}
+```
+
+Multiple image editing:
+```json
+{
+  "prompt": "Apply a consistent cyberpunk aesthetic to all images with neon lights and futuristic elements",
+  "image_urls": [
+    "https://example.com/character1.jpg",
+    "https://example.com/character2.jpg",
+    "https://example.com/background.jpg"
+  ],
+  "image_resolution": "4K",
+  "max_images": 3
+}
+```
+
+**Key Features:**
+- **Unified Interface**: Single tool for both text-to-image and image editing
+- **Smart Mode Detection**: Automatically detects mode based on presence of `image_urls`
+- **High Resolution**: Support for 1K, 2K, and 4K output
+- **Multiple Images**: Generate up to 6 images in a single request
+- **Batch Editing**: Edit up to 10 images simultaneously with consistent style
+- **Reproducible Results**: Seed control for consistent output
+
+**Note**: The `callBackUrl` is optional and will use the `KIE_AI_CALLBACK_URL` environment variable if not provided. Image generation typically takes 30-120 seconds depending on resolution and complexity.
+
+### 14. `runway_aleph_video`
+Transform videos using Runway Aleph video-to-video generation with AI-powered editing.
+
+**Parameters:**
+- `prompt` (string, required): Text prompt describing desired video transformation (max 1000 chars)
+- `videoUrl` (string, required): URL of the input video to transform
+- `waterMark` (string, optional): Watermark text to add to the video (max 100 chars, default: "")
+- `uploadCn` (boolean, optional): Whether to upload to China servers (default: false)
+- `aspectRatio` (enum, optional): Output video aspect ratio (default: "16:9")
+  - Options: `16:9`, `9:16`, `4:3`, `3:4`, `1:1`, `21:9`
+- `seed` (integer, optional): Random seed for reproducible results (1-999999)
+- `referenceImage` (string, optional): URL of reference image for style guidance
+- `callBackUrl` (string, optional): URL for task completion notifications
+
+**Examples:**
+
+Basic video transformation:
+```json
+{
+  "prompt": "Transform this video into a cinematic anime style with vibrant colors",
+  "videoUrl": "https://example.com/input-video.mp4",
+  "aspectRatio": "16:9"
+}
+```
+
+Advanced transformation with reference image:
+```json
+{
+  "prompt": "Apply the artistic style of the reference image to this video",
+  "videoUrl": "https://example.com/cooking-video.mp4",
+  "referenceImage": "https://example.com/van-gogh-painting.jpg",
+  "seed": 123456,
+  "waterMark": "My Channel"
+}
+```
+
+Vertical video for social media:
+```json
+{
+  "prompt": "Convert to a dreamy, ethereal style with soft lighting",
+  "videoUrl": "https://example.com/landscape-video.mp4",
+  "aspectRatio": "9:16",
+  "uploadCn": false
+}
+```
+
+**Key Features:**
+- **Video-to-Video Transformation**: Transform existing videos with AI-powered editing
+- **Style Transfer**: Apply artistic styles from text prompts or reference images
+- **Aspect Ratio Control**: Convert between horizontal, vertical, and square formats
+- **Reproducible Results**: Seed control for consistent transformations
+- **Watermark Support**: Add custom watermarks to transformed videos
+- **Reference Guidance**: Use reference images to guide the transformation style
+
+**Note**: The `callBackUrl` is optional and will use the `KIE_AI_CALLBACK_URL` environment variable if not provided. Video-to-video transformation typically takes 3-8 minutes depending on complexity and length.
+
+### 14. `wan_video`
+Generate videos using Alibaba Wan 2.5 models (unified tool for both text-to-video and image-to-video).
+
+**Parameters:**
+- `prompt` (string, required): Text prompt for video generation (max 800 chars)
+- `image_url` (string, optional): URL of input image for image-to-video generation (if not provided, uses text-to-video)
+- `aspect_ratio` (string, optional): Video aspect ratio for text-to-video (default: "16:9")
+  - Options: `16:9`, `9:16`, `1:1`
+- `resolution` (string, optional): Video resolution (default: "1080p")
+  - `720p`: Faster generation
+  - `1080p`: Higher quality
+- `duration` (string, optional): Video duration for image-to-video (default: "5")
+  - Options: `5`, `10` seconds
+- `negative_prompt` (string, optional): Negative prompt to describe content to avoid (max 500 chars, default: "")
+- `enable_prompt_expansion` (boolean, optional): Enable prompt rewriting using LLM (default: true)
+- `seed` (integer, optional): Random seed for reproducible results
+- `callBackUrl` (string, optional): URL for task completion notifications
+
+**Examples:**
+
+Text-to-video generation:
+```json
+{
+  "prompt": "A dimly lit jazz bar at night, wooden tables glowing under warm pendant lights. Patrons sip drinks and chat quietly while a three-piece band performs on stage. The saxophone player stands under a spotlight, gleaming instrument reflecting the light. No dialogue. Ambient audio: smooth live jazz music with saxophone and piano, clinking glasses, low murmur of audience conversations.",
+  "aspect_ratio": "16:9",
+  "resolution": "1080p",
+  "enable_prompt_expansion": true,
+  "seed": 42
+}
+```
+
+Image-to-video generation:
+```json
+{
+  "prompt": "The same woman from the reference image looks directly into the camera, takes a breath, then smiles brightly and speaks with enthusiasm: 'Have you heard? Alibaba Wan 2.5 API is now available on Kie.ai!'",
+  "image_url": "https://example.com/portrait.jpg",
+  "duration": "5",
+  "resolution": "1080p",
+  "negative_prompt": "blurry, low quality",
+  "seed": 123
+}
+```
+
+**Key Features:**
+- **Unified Interface**: Single tool for both text-to-video and image-to-video
+- **Smart Mode Detection**: Automatically detects mode based on presence of `image_url`
+- **Prompt Expansion**: LLM-powered prompt rewriting for better results with short prompts
+- **Flexible Resolutions**: 720p for speed, 1080p for quality
+- **Aspect Ratio Control**: Support for horizontal, vertical, and square formats (text-to-video)
+- **Duration Control**: 5 or 10 second options for image-to-video
+- **Negative Prompts**: Fine-tune results by specifying what to avoid
+- **Reproducible Results**: Seed control for consistent output
+
+**Note**: The `callBackUrl` is optional and will use the `KIE_AI_CALLBACK_URL` environment variable if not provided. Video generation typically takes 2-6 minutes depending on resolution and complexity.
+
 ## Why Developers Choose Kie.ai Over Alternatives
 
 ### 💸 **Better Value Than Fal.ai**
@@ -834,7 +961,7 @@ nano_banana_generate: "Modern minimalist app icon for fitness tracker"
 bytedance_seedance_video: "Screen recording showing app features, clean interface"
 
 # Add narration
-elevenlabs_tts_turbo: "Tap here to get started with your new profile"
+elevenlabs_tts: "Tap here to get started with your new profile"
 ```
 
 ### 🏢 **Enterprise Applications**