Image generation

OpenAI-compatible image generation with per-image pricing.

Image models use the OpenAI Images API surface — the same one client.images.generate speaks in every OpenAI SDK:

curl -s https://api.gopuram.net/v1/images/generations \
  -H "Authorization: Bearer $GOPURAM_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "bytedance/seedream-4.0",
    "prompt": "a majestic gopuram temple tower at golden hour",
    "size": "1024x1024"
  }' | jq -r '.data[0].b64_json' | base64 -d > out.png

Or with the SDK:

const image = await client.images.generate({
  model: 'bytedance/seedream-4.0',
  prompt: 'a majestic gopuram temple tower at golden hour',
  size: '1024x1024',
});
// image.data[0].b64_json

Request

Field	Meaning
`model`	an image model id from the catalog (`output_modalities: ["image"]`)
`prompt`	up to 10,000 characters
`n`	1–4 images
`size`	`WIDTHxHEIGHT`, e.g. `1024x1024` (model-dependent)
`response_format`	`b64_json` (default). Hosted URLs are coming with artifact storage.

Response

{
  "created": 1787443200,
  "data": [{ "b64_json": "…" }],
  "usage": { "images": 1, "cost": 0.0315 }
}

As everywhere on Gopuram, the actual cost is in-band — image models are priced per generated image (see the pricing.image field on /v1/models), so the charge is exact and known even before generation starts.

Remixing with reference images

Models whose catalog entry lists image in input_modalities accept reference images through /v1/images/edits — the same multipart surface as OpenAI's client.images.edit:

curl -s https://api.gopuram.net/v1/images/edits \
  -H "Authorization: Bearer $GOPURAM_API_KEY" \
  -F "model=bfl/flux-kontext-pro" \
  -F "prompt=the same scene during a monsoon thunderstorm" \
  -F "image=@reference.png" \
  | jq -r '.data[0].b64_json' | base64 -d > remixed.png

const remixed = await client.images.edit({
  model: 'bfl/flux-kontext-pro',
  prompt: 'the same scene during a monsoon thunderstorm',
  image: fs.createReadStream('reference.png'),
});

Pass several references by repeating the image field (up to 8 MB per file). Pricing is identical to generation: per output image, exact, in-band. Sending a reference to a model that doesn't accept one returns a 400 that says so.

Current image models

Model	Per image	References	Best at
`google/gemini-3.1-flash-image`	$0.07035	yes	Nano Banana 2 — instruction-following edits
`google/gemini-3.1-flash-lite-image`	$0.0357	yes	Nano Banana Lite — fast, cheap remixing
`google/gemini-3-pro-image`	$0.14112	yes	Nano Banana Pro — highest fidelity
`google/gemini-2.5-flash-image`	$0.04095	yes	the original Nano Banana
`bytedance/seedream-5.0-pro`	$0.03675	yes	current ByteDance flagship
`bytedance/seedream-4.5`	$0.042	yes	strong all-rounder
`bytedance/seedream-4.0`	$0.0315	yes	the price floor
`spacexai/grok-imagine-image-2.0`	$0.063	yes	xAI's latest
`recraft/recraft-v4.1`	$0.03675	no	design, brand and vector-style output
`bfl/flux-kontext-pro`	$0.042	yes	faithful remixing — preserves source structure
`bfl/flux-pro-1.1`	$0.042	no	classic high-quality text-to-image

The openai/gpt-image-1.5 family is also available — it bills by token rather than per image, so its exact cost appears in usage.cost per request. Reference support is probe-verified per model, not assumed. Prices include margin and match the in-band usage.cost exactly. The authoritative list is always the live catalog:

Which endpoint for which model

Chat models belong to /v1/chat/completions, image models to /v1/images/generations — using the wrong endpoint returns a 400 that points you to the right one. Filter the catalog programmatically:

curl -s https://api.gopuram.net/v1/models \
  | jq '.data[] | select(.architecture.output_modalities == ["image"]) | {id, pricing}'