Gemini 3.1 Flash Image Preview

• Model: gemini-3.1-flash-image-preview (fixed)
• Features: Text-to-image, image-to-image (up to 14 references), up to 4K, extreme aspect ratios, optional Google search grounding

Do not call this model via /v1/chat/completions — you will get 400 with a hint to use the Images API.

Endpoints

Mode	Endpoint	When to use
Sync (recommended)	POST https://apimaster.ai/v1/images/generations	One HTTP round-trip; platform polls upstream and returns { "data": [{ "url" }] }
Async	POST https://apimaster.ai/v1/images/generations/async → GET https://apimaster.ai/v1/tasks/{task_id}?model=gemini-3.1-flash-image-preview	Long jobs, custom queues

Quick start (sync)

curl -s "https://apimaster.ai/v1/images/generations" \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "gemini-3.1-flash-image-preview",
    "prompt": "Cyberpunk city at night, neon lights",
    "size": "16:9",
    "resolution": "2K",
    "n": 1
  }'

Success

{
  "created": 1782633696,
  "data": [{ "url": "https://apimaster.ai/imgs/1782633695067198543.jpg" }]
}

Set HTTP read timeout to ≥ 180s for 2K/4K. Download or mirror URLs promptly — they expire.

Async flow

Submit

curl -s "https://apimaster.ai/v1/images/generations/async" \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "gemini-3.1-flash-image-preview",
    "prompt": "a corgi astronaut on the moon",
    "size": "1:1",
    "resolution": "1K",
    "n": 1
  }'

Submit response

{
  "code": 200,
  "data": [{ "status": "submitted", "task_id": "task_01K8SGYNNNVBQTXNR4MM964S7K" }]
}

Poll

curl -s "https://apimaster.ai/v1/tasks/TASK_ID?model=gemini-3.1-flash-image-preview" \
  -H "Authorization: Bearer YOUR_API_KEY"

On success read data.result.images[0].url. Typical poll statuses: in_progress → succeeded / completed.

Wait 10–20s before the first poll, then every 3–5s.

Auth

Authorization: Bearer YOUR_API_KEY
Content-Type: application/json

Body parameters

Field	Type	Required	Notes
model	string	yes	gemini-3.1-flash-image-preview
prompt	string	yes	Scene description
size	string	no	Aspect ratio; see below
resolution	string	no	0.5K / 1K / 2K / 4K, default 1K; case-insensitive
n	integer	no	Only 1 supported (numeric, not string)
image_urls	array	no	Reference images (URL or data URI)
google_search	boolean	no	Google text search grounding
google_image_search	boolean	no	Requires google_search: true

Resolution & billing tiers

Value	~pixels	Price vs 1K base
0.5K	~512	same as 1K
1K	~1024	base
2K	~2048	× 4/3
4K	~4096	× 2

Actual charge = channel price × resolution multiplier × group ratio. See marketplace or console logs.

Errors

HTTP	Meaning
400	Bad params, or wrong endpoint (chat)
401	Invalid API key
402	Insufficient balance
408	Sync poll timeout — retry async or lower resolution
500 / 502	Upstream or platform error

Invalid aspect ratios may surface as 500 when only one channel is configured.

Notes

• Usage logs include result_url for image preview and request_data.effective_resolution.
• Pricing varies by channel and resolution — see marketplace.