Image generation

ru-llm exposes image models (gpt-image-2, gemini-3.1-flash-image) through an OpenAI-compatible image-generation endpoint.

Endpoint

POST https://api.ru-llm.relay2.xyz/v1/images/generations

Request

curl https://api.ru-llm.relay2.xyz/v1/images/generations \
  -H "Authorization: Bearer sk-ru-YOUR_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "gpt-image-2",
    "prompt": "A red panda coding at a laptop, soft studio lighting",
    "size": "1024x1024",
    "n": 1
  }'

Field	Type	Required	Description
`model`	string	Yes	Image model ID (e.g. `gpt-image-2`, `gemini-3.1-flash-image`).
`prompt`	string	Yes	Text description of the image to generate.
`size`	string	No	Output resolution, e.g. `1024x1024`.
`n`	integer	No	Number of images to generate. Default: 1.

Response

The response follows the OpenAI images format, returning base64-encoded image data:

{
  "created": 1710000000,
  "data": [
    {"b64_json": "iVBORw0KGgoAAAANSUhEUgAA..."}
  ]
}

Decode the first image to a PNG file:

curl ... | python3 -c "import sys,json,base64; \
open('out.png','wb').write(base64.b64decode(json.load(sys.stdin)['data'][0]['b64_json']))"

Billing

Billing depends on the model. gpt-image-2 is billed per token (input/output), the same way chat models are. gemini-3.1-flash-image is billed per generated image (the per_image rate), charged once the image is produced. See the Models & Pricing page for current rates.

Troubleshooting

Symptom	Cause	Fix
`404 Not Found`	Endpoint not yet mounted (see preview note)	Use chat/messages until the image surface ships.
`401 Unauthorized`	Key missing or wrong	Pass a valid `sk-ru-…` key.
`402 Payment Required`	Zero balance	Top up in the console → Billing.
`400` unknown model	Model ID not an image model	Use `gpt-image-2` or `gemini-3.1-flash-image`.