Flux Image Generation

Overview

• Method: POST
• Path: /v1/images/generations
• Content-Type: application/json
• Description: This request format is consistent with OpenAI DALL-E.

Supported sizes and ratios

Size	Ratio
1024x1024	1:1
512x1024	1:2
1024x512	2:1
768x512	3:2
512x768	2:3
960x1280	3:4
1280x960	4:3
720x1280	9:16
1280x720	16:9
576x1344	9:21
1344x576	21:9

---

Authentication

• Authorization: Bearer sk-xxx (required)

---

Request Example

Header

Header	Example	Required
Content-Type	application/json	Yes
Authorization	Bearer sk-xxxx	Yes

Request Body (JSON)

{
  "prompt": "A detailed image describing a vast Stone Age city...",
  "model": "flux-kontext-pro",
  "size": "3:4",
  "output_format": "png",
  "seed": 12345,
  "prompt_upsampling": true,
  "safety_tolerance": 6
}

Parameter Description

Parameter	Type	Required	Default	Description
prompt	string	Yes	-	Text description of the image, max 1000 (or 4000) chars
model	string	Yes	-	Model name; supports flux, flux-dev, flux-pro, flux-pro-max, flux.1.1-pro, flux-kontext-max, flux-kontext-pro
size	string	No	1:1	Image size, supports related ratios or numeric formats (width/height multiples of 32)
output_format	string	No	png	jpeg or png
seed	integer	No	-	Random seed
prompt_upsampling	boolean	No	true	Enable prompt optimization
safety_tolerance	integer	No	6	Safety tolerance 0-6 (0 strictest, 6 lenient)

---

Response Example (200)

{
  "created": 1725007743,
  "data": [
    {
      "revised_prompt": "A detailed image illustrating a sprawling Stone Age city...",
      "url": "https://oaidalleapiprodscus.blob.core.windows.net/private/org-ckj8cP1MmQK3ZdLPGoozEzD9/vv/img-l4C07hxN49jraolwGOXGc6Ce.png?..."
    }
  ]
}