Two pricing tracks
At the core, GPT Image 2 is priced as an API model. The named API model is gpt-image-2, while chatgpt-image-latest is the moving alias that tracks the snapshot currently used inside ChatGPT. Separately, OpenAI also gives end users access to the broader image stack through ChatGPT plans, which is the consumer surface many readers call ChatGPT Images 2.0 or ImageGen 2.0 in ChatGPT.
ChatGPT plan pricing
For most solo creators, Go is the cheapest paid entry point and Plus is the cleaner serious-creator tier. Pro only makes sense if image generation is a daily, high-volume part of your work. Business matters once you need shared workspaces, admin controls, or procurement-friendly billing.
API pricing for gpt-image-2
OpenAI prices GPT Image 2 as usage, not a flat subscription. The official API pricing page lists text and image token rates, while the image generation guide publishes approximate common-size output costs.
GPT Image 2 also supports flexible sizes, with popular presets reaching 2K and 4K. That makes the API track better suited to product features, batch creative operations, or teams that want fine control over size, quality, and editing inputs.
What about chatgpt-image-latest?
chatgpt-image-latest is the API alias that tracks the image snapshot used inside ChatGPT. OpenAI currently lists it separately from gpt-image-2, so developers who want the newest flagship API model should price gpt-image-2, while teams trying to mirror ChatGPT image behavior should review the chatgpt-image-latest model page before shipping.
Best value by use case
Use ChatGPT Go or Plus if you mainly create images for yourself, your team, or client work inside a chat interface.
Use the API if you need product integration, predictable rendering settings, automated edits, or large-volume generation. That is where GPT Image 2's newer API pricing and flexible sizes matter most.