Image prompt
Optional reference images in Vizify
Size and quality
Format and delivery settings
GPT Image 2 tool
Create images with GPT Image 2.
Describe the image you need, choose only the output settings that matter, and carry a model-ready request into Vizify.
GPT Image 2 with automatic provider routing Only supported output controls Prompt and settings stay together
Start with the messy versionOptional choices help Vizify shape the handoff.
Try a starting point
Vizify prefers the Wuyin GPT Image 2 route. Choosing an exact size, quality, format, compression, or background requirement switches the request to the OpenAI route.
GPT Image 2 generation request
One or more generated images
Reusable prompt and settings
What it does
Built to hand back finished work.
Flexible resolution without guesswork
Start with Auto or choose common square, portrait, landscape, 2K, and 4K presets that follow GPT Image 2 resolution constraints.
Draft quickly, finish deliberately
Use Low for faster exploration, then move to Medium or High when the composition is ready for final rendering.
Reference images stay high fidelity
When you continue in Vizify with reference images, GPT Image 2 preserves their details at high fidelity automatically—there is no fidelity toggle to manage.
Delivery settings that match the asset
Keep output delivery on Auto for the default Wuyin route, or choose PNG, JPEG, or WebP to use OpenAI. Vizify can fan out up to ten independent image jobs in one run.
FAQ
Questions, answered.
Which inputs can I control?
The useful user-facing inputs are the prompt, optional reference images, image size, quality, output format, JPEG/WebP compression, background behavior, and number of images. Vizify keeps GPT Image 2 fixed while selecting its Wuyin or OpenAI route according to the requested controls.
Can I generate a transparent background?
Not with GPT Image 2 currently. The supported background choices are Auto and Opaque, so this tool does not show a transparent option.
Why is there no input fidelity setting?
GPT Image 2 processes every reference image at high fidelity automatically. The API does not allow that behavior to be changed, so exposing a fidelity selector would be misleading.
Can I enter a custom resolution?
GPT Image 2 supports custom WIDTHxHEIGHT resolutions when both edges are multiples of 16, the aspect ratio stays between 1:3 and 3:1, the longest edge is at most 3840 px, and the total pixel count stays within the model limits. Vizify can validate a custom size before generation.
What does GPT Image 2 Generator deliver?
The workflow returns GPT Image 2 generation request, One or more generated images, Reusable prompt and settings. The exact artifact remains available for review in Vizify.
Which parts of the GPT Image 2 Generator result can I customize?
You can revise the prompt, selected controls, required content, composition direction, and supported delivery settings before or after the first run in Vizify.
Can I use reference images with GPT Image 2?
Yes. Continue in Vizify and attach reference images when you need editing, subject consistency, or visual direction. The Wuyin GPT Image 2 route accepts up to eight reference image URLs and processes them at high fidelity.
How many GPT Image 2 images can I generate at once?
You can request from one to ten images. Vizify starts an independent generation job for each image, so the count works without relying on an unsupported provider-level count parameter.
GPT Image 2 input parameters
This tool keeps the creative controls visible and leaves infrastructure settings to Vizify. The prompt is required. Reference images are optional and are used for editing, subject consistency, or visual direction. Size, quality, output format, background, compression, and image count are optional.
Visible controls
- Prompt: subject, composition, environment, lighting, visual style, color, camera or medium, and exact text requirements.
- Reference images: optional source material for edits or for preserving a subject, product, or visual direction.
- Size: Auto, common presets, or a validated custom resolution.
- Quality: Auto, Low, Medium, or High.
- Output format: PNG, JPEG, or WebP.
- Compression: 0–100 for JPEG and WebP only.
- Background: Auto or Opaque. GPT Image 2 does not currently support Transparent.
- Image count: 1–10 outputs per request.
Handled automatically by Vizify
The model family is fixed to GPT Image 2. Vizify uses gpt-image-2-wuyin by default for prompts, aspect ratios, and reference images. An exact pixel size, non-Auto quality, explicit output format, compression, background, or moderation requirement selects gpt-image-2-openai. Image count is implemented by Vizify as independent jobs rather than passed as an unsupported provider parameter. Reference images use the Wuyin route; moderation, streaming, partial-image delivery, and the end-user identifier remain runtime concerns.
Prompt, settings, finished asset
Open Vizify