Update Logs

Welcome to the beta of gptimage2.design — a creative workspace built around OpenAI's GPT Image 2 model, live today in our generator, plus the current generation of top competing image models in the same canvas.

This is a soft launch. Expect rough edges; file bugs at support@gptimage2.design.

What's included in v0.1

• Text-to-image canvas with GPT Image 1.5, Nano Banana Pro, and MAI-Image-2 as available backends
• Image-to-image with single and multi-reference support (up to 4 references per request)
• Aspect ratio picker — 1:1, 16:9, 9:16, 4:3, 3:4, custom
• Quality tiers — Draft / Standard / HD per generation
• Credit-based billing with one-time top-ups and monthly subscription plans
• Account, billing, and activity dashboard
• Prompt history with quick "rerun" on any past generation
• Dark mode and English / 简体中文 i18n

Now live

• GPT Image 2 — live in the generator; select it in the model dropdown. We route to OpenAI's official public API automatically once they ship it.
• Prompt library — 18 curated GPT Image 2 prompts with verified outputs at /gpt-image-2-prompts

What's coming before the v1.0 release

• Character lock across batches
• Region-based prompting UI
• Team workspaces with shared asset libraries

Notes on pricing

• Early-bird credits for the first 1,000 signups
• Annual plan unused-credit rollover
• Per-generation cost scales with quality tier and resolution

We'll post a changelog entry every week. Subscribe to updates on the changelog page or via our mailing list.