Model Catalog

Pixapi exposes multiple model families behind one API key. Pick a model id, pass it in the OpenAI-compatible request body, and keep the rest of your integration stable while you compare providers.

Image models

Image generation uses POST /v1/images/generations. Image editing uses POST /v1/images/edits when the model supports input images. Use n, size, and optional controls such as quality. For large outputs or slower models, submit through /v1/async/images/generations or /v1/async/images/edits, then poll GET /v1/tasks/{id}.

Model	Model id	Best for	Endpoints	Credits
Gemini 2.5 Flash Image	gemini-2.5-flash-image (nano-banana alias)	Fast Gemini image generation and editing	Generate, edit	2
Gemini 3.1 Flash Image Preview	gemini-3.1-flash-image-preview (nano-banana-2 alias)	Size-controlled 512, standard, 2K, and 4K outputs	Generate, edit	from 4
Gemini 3 Pro Image Preview	gemini-3-pro-image-preview (nano-banana-pro alias)	Premium generation and edits with standard, 2K and 4K outputs controlled by size	Generate, edit	from 8
GPT Image 1.5	gpt-image-1.5	GPT image generation and editing with strong text rendering	Generate, edit	from 1
GPT Image 2	gpt-image-2	OpenAI quality tiers, custom sizes, prompt adherence	Generate, edit	from 1

Video models

Model	Model id	Best for	Endpoint	Credits
Veo	veo	Realistic motion and cinematic product clips	/v1/async/videos/generations	from 1/sec
Kling	kling	Expressive motion, subject control, and creative product clips	/v1/async/videos/generations	from 2.24/sec
Wan	wan	Creative short video generation	/v1/async/videos/generations	from 2/sec

Credit deduction

Gemini image requests are billed by selected model and output resolution.

Model id	Credits / image
gemini-2.5-flash-image	2
gemini-3.1-flash-image-preview-512	4
gemini-3.1-flash-image-preview	4
gemini-3.1-flash-image-preview-2k	6
gemini-3.1-flash-image-preview-4k	8
gemini-3-pro-image-preview	8
gemini-3-pro-image-preview-2k	8
gemini-3-pro-image-preview-4k	16

GPT Image 2 is billed by output size tier and quality. quality=auto is billed as high, and size=auto defaults to the 1K tier.

Size tier	Long edge	low	medium	high / auto
1K	<= 1536px	1	4	14
2K	<= 2048px	2	17	67
4K	<= 3840px	4	34	133

Video models are billed by successful generated duration. Prices below are 20% of each provider's official latest model price, with 1 credit = $0.01.

Video generation is async-only. Use /v1/async/videos/generations and the Async Media Tasks flow for all video integrations.

Model / tier	Official price	Pixapi credits
Veo 3.1 Lite 720P	$0.05/sec	1/sec
Veo 3.1 Lite 1080P	$0.08/sec	1.6/sec
Veo 3.1 Fast 720P	$0.10/sec	2/sec
Veo 3.1 Fast 1080P	$0.12/sec	2.4/sec
Veo 3.1 Fast 4K	$0.30/sec	6/sec
Veo 3.1 Standard 720P/1080P	$0.40/sec	8/sec
Veo 3.1 Standard 4K	$0.60/sec	12/sec
Kling 3.0 Turbo 720P	$0.112/sec	2.24/sec
Wan2.7 720P	$0.10/sec	2/sec
Wan2.7 1080P	$0.15/sec	3/sec

Choosing a model

• Use gemini-2.5-flash-image for quick image output and editing workflows. nano-banana is an alias for the same model.
• Use gemini-3.1-flash-image-preview when you need a balanced Gemini preview model with broad size support. nano-banana-2 is an alias for the same model.
• Use gemini-3-pro-image-preview when quality matters more than cost. nano-banana-pro is an alias for the same model.
• Use gpt-image-2 when prompt adherence, typography, custom size, or layout precision matters.
• Use veo when the output needs motion, camera direction, and video timing.

Stable integration pattern

The endpoint usually stays the same for a media type. Switching models is normally a request-body change:

{
  "model": "gemini-3-pro-image-preview",
  "prompt": "A clean product hero image for a developer API platform",
  "n": 1,
  "size": "1024x1024"
}

Legacy Nano Banana-compatible fields are separated under Legacy transition format.