Image Edition

Edit an Image

POST/v1/images/edits

Edit an image following instructions given by the prompt. Compatible with OpenAI Image Edition API with some limitations (cf column "support" in the Request Body table).

---

Request

curl -X POST https://api.ig1.ai/v1/images/edits \
  -H "Authorization: Bearer $IG1AI_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "qwen-image-edit",
    "images": [
      {
        "image_url": "data:image/png;base64,<BASE64_ENCODED_IMAGE_DATA>"
      }
    ],
    "prompt": "Replace the iguana with a snake, keep the same neon theme and background.",
    "quality": "low",
    "size": "1280x720",
    "output_format": "png"
  }'

Request body

Parameter	Type	Required	Support	Description
prompt	string	✅	✅ Full	Edit instruction
model	string	✅	✅ Full	qwen-image-edit or qwen-image-edit-pe
images	array	✅	⚠️ Partial	Array of image objects. Only image_url with base64 data URL supported. External URLs not supported
output_format	string		✅ Full	png, jpeg, or webp
quality	string		✅ Full	low, medium, standard, high, hd, auto
size	string		✅ Full	Target output size
n	integer		⚠️ Partial	Only 1 accepted
response_format	string		⚠️ Partial	Only b64_json supported
background			❌ Ignored
input_fidelity			❌ Ignored
mask			❌ Ignored
moderation			❌ Ignored
output_compression			❌ Ignored
partial_images			❌ Ignored
stream			❌ Ignored
user			❌ Ignored

---

Response

{
  "created": 1773764659,
  "data": [
    {
      "b64_json": "<BASE64_ENCODED_IMAGE_DATA>"
    }
  ],
  "usage": {
    "total_tokens": 0,
    "input_tokens": 0,
    "output_tokens": 36864000,
    "input_tokens_details": {
      "text_tokens": 0,
      "image_tokens": 0
    }
  },
  "output_format": "png",
  "quality": "low",
  "size": "1280x720"
}

Prompt Enhancement

Use model qwen-image-edit-pe to automatically refine your prompt via Qwen3.5-397B-A17B before editing. Prompt enhancement override your prompt with more details and elements that fit perfectly with model's expectations to generate a better image.

info

Prompt enhancement tokens are charged under the Qwen3.5-397B-A17B model.

Unsupported Open AI endpoints

warning

POST/v1/images/variations endpoint is not supported by IG1 AI API.