GPT Image 2 Serverless API

Generate photorealistic images with legible multilingual text and 2K output.

~104.42s
POST /v2/gpt-image-2 · submit + poll
 1# pip install "segmind>=1.1.0"
 2# export SEGMIND_API_KEY="YOUR_API_KEY"
 3import segmind
 4
 5# Async (v2): submit to the queue and block until COMPLETED.
 6# run() returns the final result dict (600s deadline, 1.0s poll by default).
 7result = segmind.run(
 8    "gpt-image-2",
 9    size="1536x1024",
10    prompt="A photorealistic, cinematic shot of a cozy independent bookstore in Mumbai at golden hour. Warm afternoon sunlight streams through a tall front window onto wooden shelves packed with books; book spines are clearly visible with titles in English and Hindi Devanagari script. In the foreground, a handwritten chalkboard A-frame easel reads, in clean legible chalk lettering: first line 'मुंबई पुस्तक भंडार', second line 'Mumbai Book Store', and a smaller third line 'Open Daily  9 am – 9 pm'. Shallow depth of field, shot on a full-frame camera at 35mm f/2.0, ultra-realistic detail, natural color grading, dust motes in sunlight, film grain.",
11    quality="high",
12    background="opaque",
13    image_urls=[],
14    moderation="auto",
15    output_format="png",
16    output_compression=100,
17)
18print(result["status"])                      # COMPLETED
19print(result.get("output"))                  # model output (e.g. media URL)
20print(result["metrics"]["inference_time"])   # server compute seconds
21
22# --- Or submit + poll manually (track request_id, control the cadence) ---
23from segmind import SegmindClient, InferenceFailed, InferenceTimeout
24
25client = SegmindClient()                      # reads SEGMIND_API_KEY
26payload = {
27    "size": "1536x1024",
28    "prompt": "A photorealistic, cinematic shot of a cozy independent bookstore in Mumbai at golden hour. Warm afternoon sunlight streams through a tall front window onto wooden shelves packed with books; book spines are clearly visible with titles in English and Hindi Devanagari script. In the foreground, a handwritten chalkboard A-frame easel reads, in clean legible chalk lettering: first line 'मुंबई पुस्तक भंडार', second line 'Mumbai Book Store', and a smaller third line 'Open Daily  9 am – 9 pm'. Shallow depth of field, shot on a full-frame camera at 35mm f/2.0, ultra-realistic detail, natural color grading, dust motes in sunlight, film grain.",
29    "quality": "high",
30    "background": "opaque",
31    "image_urls": [],
32    "moderation": "auto",
33    "output_format": "png",
34    "output_compression": 100,
35}
36job = client.submit_async("gpt-image-2", **payload)
37print(job.request_id)                         # available immediately
38try:
39    result = job.wait(timeout=600, interval=1.0)
40except InferenceTimeout as e:
41    print("still running:", e.request_id)
42except InferenceFailed as e:
43    print("failed:", e.detail)

API Endpoint

POSThttps://api.segmind.com/v1/gpt-image-2

Parameters

promptrequired
string

Text describing the image; supports in-image typography across scripts. Lead with subject, style, lighting.

backgroundoptional
string

'opaque' for full scenes; 'transparent' for logos, stickers, and product cutouts.

Default: "opaque"
Allowed values :
"opaque"
heightoptional
integer

Optional custom height in pixels. Set both Width and Height to override the Size preset. Rules: each a multiple of 16, aspect ratio between 1:3 and 3:1, longest edge <= 3840, total pixels 655,360-8,294,400.

Default: null
image_urlsoptional
string[]

A list of reference images. Include one or more URLs to edit or draw context from.

mask_image_urloptional
string (uri)

Optional mask image URL for surgical inpainting. White regions of the mask indicate areas to edit; everything outside stays pixel-perfect.

Default: ""
moderationoptional
string

Content filter strictness; 'auto' is the safe default. Use 'low' only for permitted use cases.

Default: "auto"
Allowed values :
"low""auto"
output_compressionoptional
integer

Compression level 0-100; 100 preserves text crispness. Lower values reduce file size.

Default: 100
output_formatoptional
string

Use 'png' for crisp text, 'webp' for size, 'jpeg' for broad compatibility.

Default: "png"
Allowed values :
"png""jpeg""webp"
qualityoptional
string

Rendering fidelity; 'high' keeps typography crisp. Use 'medium' or 'low' only for previews.

Default: "high"
Allowed values :
"low""medium""high""auto"
sizeoptional
string

Output resolution (WIDTHxHEIGHT). 'auto' lets the model pick. For a custom resolution not listed, set Width and Height instead. Constraints: each edge a multiple of 16, aspect ratio 1:3-3:1, longest edge <= 3840, total pixels 655,360-8,294,400.

Default: "1536x1024"
Allowed values (13 total):
"1024x1024""1536x1024""1024x1536""1280x960""960x1280""1536x864""864x1536""2048x2048""2048x1152""1152x2048"+3 more
widthoptional
integer

Optional custom width in pixels. Set both Width and Height to override the Size preset. Rules: each a multiple of 16, aspect ratio between 1:3 and 3:1, longest edge <= 3840, total pixels 655,360-8,294,400.

Default: null

Response Type

Returns: Text/JSON

Asynchronous requests (v2)

Use Async for video, long-running (>~60s), or high-concurrency workloads; Sync is simplest for fast image & LLM calls. Async submits a request and you poll it to completion.

  1. 1
    POST /v2/gpt-image-2

    Submitreturns request_id, status_url, response_url

  2. 2
    GET /v2/requests/{id}/status

    Polluntil COMPLETED or FAILED

  3. 3
    GET /v2/requests/{id}

    Resultfinal response body

Status states

QUEUEDAccepted, waiting for a worker
PROCESSINGRunning on a worker
COMPLETEDDone — result body is ready
FAILEDErrored (incl. content/RAI blocks)
  • A FAILED request is served as HTTP 422 — the body still carries the error detail.
  • An unknown or expired request_id returns HTTP 404.
  • Results are retained for 1 hour, then expire.
  • Content / RAI blocks surface as FAILED, not a separate state.
  • Track completion by polling the status endpoint.

Common Error Codes

The API returns standard HTTP status codes. Detailed error messages are provided in the response body.

400

Bad Request

Invalid parameters or request format

401

Unauthorized

Missing or invalid API key

403

Forbidden

Insufficient permissions

404

Not Found

Model or endpoint not found

406

Insufficient Credits

Not enough credits to process request

429

Rate Limited

Too many requests

500

Server Error

Internal server error

502

Bad Gateway

Service temporarily unavailable

504

Timeout

Request timed out