GPT Image 2
GPT Image 2 is OpenAI's versatile image model, with excellent instruction-following and among the best legible in-image text. A flexible all-rounder for graphics, layouts and product scenes.
From 2–42 credits per image, included in every plan.
Real generations
Every example here was genuinely made with this model in Oxava — no stock.
What it's good at
How to use it in Oxava
Pick GPT Image 2
Choose the model and, if editing, add your reference image.
Write clear instructions
It follows detailed prompts well — spell out text, layout and composition.
Generate & export
Pick the best variation and export or upscale as needed.
Frequently asked questions
Is GPT Image 2 good at text in images?+
Yes — it's one of the stronger models for legible in-image text, which makes it useful for social graphics, ads and packaging mockups.
What is it best for?+
Versatile design and product work, especially anything with text, precise instructions or layout requirements.
Can it edit an existing image?+
Yes — provide a reference and describe the change to transform or extend it.
Other models
Try GPT Image 2 →
GPT Image 2 is OpenAI's versatile image model, with excellent instruction-following and among the best legible in-image text. A flexible all-rounder for graphics, layouts and product scenes.