GPT-image-2 generations now work for sceneries and AI models

The OpenAI GPT-image-2 model can now generate sceneries, AI models, and reference-less photo shoots that previously failed instantly.

Selecting "Qamera AI GPT" (the OpenAI GPT-image-2 model) for a scenery, AI model, or photo shoot without a reference image now produces a result instead of failing with an "at least one reference image required" error. The provider keeps using the existing reference-based path for packshots and session photography flows. No setting changes are required from your side; existing generations and credit costs are unaffected.