Overview

We provide gpt-image-1 & gpt-image-1.5 & gpt-image-2-preview api

Usage & Pricing

gpt-image-2-preview

This is a preview version model, each generation costs $0.1

Base Token Rates

• $5.00 / 1M input text tokens
• $10.00 / 1M input image tokens
• $40.00 / 1M output image tokens

Estimated Cost per Image

To make it easier, here are the estimated prices for generating new images at common resolutions and quality settings. These costs apply to output images.

Resolution	Low Quality	Medium Quality	High Quality
1024×1024	~$0.011	~$0.042	~$0.167
1024×1536	~$0.016	~$0.063	~$0.25
1536×1024	~$0.016	~$0.063	~$0.25

Key Points

Higher resolutions and quality settings require more tokens, resulting in a higher cost per image.

Processing an image you provide (as input) is significantly less expensive than generating a new one.

These are estimates; the exact token count for a specific image may vary.

Simple Example

curl --location 'https://api.piapi.ai/v1/images/generations' \
--header 'Content-Type: application/json' \
--header 'Authorization: Bearer {your-api-key}' \
--data '{
    "model": "gpt-image-1",
    "prompt": "A cute baby sea otter",
    "n": 1,
    "size": "1024x1024",
    "quality": "low",
    "output_format": "jpeg"
}'

Image Edit with Reference Image

Use the /v1/images/edits endpoint to transform an existing image with a natural-language prompt. The request must be multipart/form-data — the reference image is uploaded as a file field named image.

Example (gpt-image-2-preview)

# 1. Prepare a reference image (either download one or use your own local file)
curl -sS -o /tmp/ref.png \
  "https://oss.filenest.top/uploads/1776b813-28cb-4ca7-a82f-6096f4e7c5b7.png"

# 2. Submit the edit request
curl --location 'https://api.piapi.ai/v1/images/edits' \
  --header 'Authorization: Bearer {your-api-key}' \
  --form 'model=gpt-image-2-preview' \
  --form 'prompt=Add a warm sunset background with soft bokeh lights, keep the character unchanged' \
  --form 'n=1' \
  --form 'size=1024x1024' \
  --form 'response_format=url' \
  --form 'image=@/tmp/ref.png'

Sample Response (HTTP 200)

{
  "data": [
    {
      "url": "https://imagefil.scdn.app/assets/codex/d69c5648-f0d0-4cd3-8860-0d7b1733d274.png"
    }
  ],
  "created": 1776956146,
  "usage": {
    "total_tokens": 1889,
    "input_tokens": 1124,
    "output_tokens": 765,
    "input_tokens_details": {
      "text_tokens": 19,
      "image_tokens": 1105
    }
  }
}

Verified sample roundtrip (~100 s end-to-end):

• Reference image: https://oss.filenest.top/uploads/1776b813-28cb-4ca7-a82f-6096f4e7c5b7.png
• Edited output: https://imagefil.scdn.app/assets/codex/d69c5648-f0d0-4cd3-8860-0d7b1733d274.png

Tips

• If response_format is omitted, the response comes back as b64_json. Pass response_format=url when you want a hosted link instead of inline base64.
• Processing a reference image you provide is billed at the input-image token rate ($10 / 1M tokens), which is significantly cheaper than generating one from scratch.
• Typical latency is 40–100 s depending on input image size and prompt complexity.