Bing Create

A REST API wrapper for Bing Image Creator (DALL-E, GPT-4o, MAI Image 1) built with Node.js + Express.

Requirements

• Node.js 18+
• A Microsoft account logged in to Bing Image Creator

Installation

git clone https://github.com/rynn-k/bing-create.git
cd bing-create
npm install

Configuration

Open bing.js and fill in your _U cookie:

static COOKIE_U = 'your_cookie_U_here';

How to get the _U cookie

1. Open https://www.bing.com/images/create in your browser
2. Sign in with your Microsoft account
3. Open DevTools → Application tab → Cookies → https://www.bing.com
4. Find the cookie named _U and copy its value

Running

node index.js

Server runs at http://localhost:7860

Endpoints

Method	Path	Description
GET	/	API info
GET	/health	Server & session status
GET	/options	Available models and aspects
POST	/create/image	Generate images
POST	/create/video	Generate video (async task)
GET	/task/:taskId	Get video task status

---

API Reference

GET /

Get API information and available endpoints.

curl http://localhost:7860/

Response:

{
  "success": true,
  "author": "rynn-k (Randyyyyy)",
  "repository": "https://github.com/rynn-k/Bing-Create",
  "endpoints": {
    "info": "GET /",
    "health": "GET /health",
    "create_image": "POST /create/image",
    "create_video": "POST /create/video",
    "get_task": "GET /task/:taskId",
    "options": "GET /options"
  }
}

---

GET /health

Check server health and session status.

curl http://localhost:7860/health

Response:

{
  "success": true,
  "data": {
    "status": "ok",
    "cookie": "set",
    "session": "active",
    "activeTasks": 2
  },
  "timestamp": "2024-01-01T00:00:00.000Z"
}

---

GET /options

Get available models and aspect ratios.

curl http://localhost:7860/options

Response:

{
  "success": true,
  "data": {
    "models": [
      { "id": "DALLE", "value": 0, "description": "DALL-E — fast, general purpose" },
      { "id": "GPT4O", "value": 1, "description": "GPT-4o — higher quality, slower" },
      { "id": "MAI1",  "value": 4, "description": "MAI Image 1 — Microsoft model" }
    ],
    "aspects": [
      { "id": "SQUARE",    "value": 1, "ratio": "1:1" },
      { "id": "LANDSCAPE", "value": 2, "ratio": "7:4" },
      { "id": "PORTRAIT",  "value": 3, "ratio": "4:7" }
    ]
  }
}

---

POST /create/image

Generate images synchronously. Returns image URLs immediately after generation completes.

Request body:

Field	Type	Default	Description
query	string	required	Prompt text
model	string|number	DALLE	DALLE, GPT4O, MAI1 or 0, 1, 4
aspect	string|number	SQUARE	SQUARE, LANDSCAPE, PORTRAIT or 1, 2, 3

Example:

curl -X POST http://localhost:7860/create/image \
  -H 'Content-Type: application/json' \
  -d '{
    "query": "a futuristic city at sunset",
    "model": "DALLE",
    "aspect": "LANDSCAPE"
  }'

Success response:

{
  "success": true,
  "data": {
    "images": [
      "https://th.bing.com/th/id/OIG...",
      "https://th.bing.com/th/id/OIG...",
      "https://th.bing.com/th/id/OIG...",
      "https://th.bing.com/th/id/OIG..."
    ],
    "prompt": "a futuristic city at sunset with neon lights and flying cars",
    "model": "DALLE",
    "aspect": "LANDSCAPE"
  },
  "timestamp": "2024-01-01T00:00:00.000Z"
}

With different model:

curl -X POST http://localhost:7860/create/image \
  -H 'Content-Type: application/json' \
  -d '{
    "query": "cyberpunk street scene",
    "model": "GPT4O",
    "aspect": "PORTRAIT"
  }'

---

POST /create/video

Generate video asynchronously. Returns a task ID to check the status later.

Request body:

Field	Type	Description
query	string	Prompt text (required)

Example:

curl -X POST http://localhost:7860/create/video \
  -H 'Content-Type: application/json' \
  -d '{
    "query": "ocean waves at sunset"
  }'

Response:

{
  "success": true,
  "data": {
    "taskId": "123e4567-e89b-12d3-a456-426614174000",
    "prompt": "ocean waves at sunset"
  },
  "timestamp": "2024-01-01T00:00:00.000Z"
}

---

GET /task/:taskId

Check the status of a video generation task.

Example:

curl http://localhost:7860/task/123e4567-e89b-12d3-a456-426614174000

Response (Processing):

{
  "success": true,
  "data": {
    "id": "123e4567-e89b-12d3-a456-426614174000",
    "status": "processing",
    "prompt": "ocean waves at sunset",
    "video": null,
    "error": null,
    "createdAt": "2024-01-01T00:00:00.000Z",
    "completedAt": null
  },
  "timestamp": "2024-01-01T00:00:05.000Z"
}

Response (Completed):

{
  "success": true,
  "data": {
    "id": "123e4567-e89b-12d3-a456-426614174000",
    "status": "completed",
    "prompt": "ocean waves at sunset",
    "video": "https://example.com/video.mp4",
    "error": null,
    "createdAt": "2024-01-01T00:00:00.000Z",
    "completedAt": "2024-01-01T00:01:30.000Z"
  },
  "timestamp": "2024-01-01T00:01:35.000Z"
}

Response (Failed):

{
  "success": true,
  "data": {
    "id": "123e4567-e89b-12d3-a456-426614174000",
    "status": "failed",
    "prompt": "ocean waves at sunset",
    "video": null,
    "error": "Prompt rejected for content policy.",
    "createdAt": "2024-01-01T00:00:00.000Z",
    "completedAt": "2024-01-01T00:00:10.000Z"
  },
  "timestamp": "2024-01-01T00:00:15.000Z"
}

Response (Not Found):

{
  "error": {
    "message": "Task not found.",
    "type": "invalid_request_error"
  }
}

---

Usage Examples

Minimal Image Generation

curl -X POST http://localhost:7860/create/image \
  -H 'Content-Type: application/json' \
  -d '{"query": "beautiful sunset"}'

High-Quality Image with GPT-4o

curl -X POST http://localhost:7860/create/image \
  -H 'Content-Type: application/json' \
  -d '{
    "query": "detailed fantasy landscape",
    "model": "GPT4O",
    "aspect": "LANDSCAPE"
  }'

Video Generation Workflow

# 1. Start video generation
TASK_ID=$(curl -s -X POST http://localhost:7860/create/video \
  -H 'Content-Type: application/json' \
  -d '{"query": "dancing flames"}' | jq -r '.data.taskId')

# 2. Check status periodically
curl http://localhost:7860/task/$TASK_ID

# 3. Keep checking until status is "completed" or "failed"

---

Error Types

Type	HTTP	Description
invalid_request_error	400	Missing or invalid parameters
auth_error	401	_U cookie is not set or has expired
prompt_rejected_error	422	Prompt violates Bing's content policy
api_error	500	Internal server error

Error Response Format

{
  "error": {
    "message": "Error description",
    "type": "error_type"
  }
}

---

Task Status Flow

POST /create/video
    ↓
  taskId returned
    ↓
GET /task/:taskId (status: "processing")
    ↓
    ... polling ...
    ↓
GET /task/:taskId (status: "completed" or "failed")

---

Notes

• Image generation is synchronous and returns results immediately
• Video generation is asynchronous and requires polling the task endpoint
• Video tasks are stored in memory and will be lost if the server restarts
• Tasks remain in memory until the server is restarted (no automatic cleanup)
• The _U cookie must be valid and have access to Bing Image Creator

---

License

MIT