GPT Image 2 AI Image Generator
Three quality tiers — scale cost to your target. Up to 4K.
Overview
GPT Image 2 is OpenAI’s image model with three quality tiers — Low, Medium and High — so you can match credit cost to how polished the result needs to be. It supports text-to-image and image editing up to 4K.
On FAV, Edit mode with auto ratio is pre-charged at 1K and adjusted to the actual output size after generation, so you never overpay.
⟳ Edit mode (auto ratio): pre-charged at 1K, adjusted to actual output size after generation.
What you can make
- ✦ Draft cheaply on Low, then finalize on High
- ✦ Edit an image while letting GPT pick the best ratio
- ✦ Generate clean, instruction-following images
GPT Image 2 FAQ
What are the quality tiers in GPT Image 2?
GPT Image 2 offers Low (1 cr), Medium (1–2 cr) and High (4–8 cr) tiers, letting you scale cost to your quality target.
How does GPT Image 2 Edit billing work?
In Edit mode with auto ratio, credits are reserved at 1K and adjusted to the actual output size after generation, so you are never charged more than the real result.
Related models
Ultra-fast text-to-image and image editing powered by Gemini.
Higher fidelity with stronger prompt adherence and richer detail.
High-quality image generation by ByteDance. Minimum 2K resolution.