ChatGPT web image generation ignores thinking mode, wasting a usage per attempt
TheSynthian · reddit · 2026-09-10
A Reddit user reports a week-long bug on ChatGPT web: when generating images with thinking enabled, the first generation never actually uses reasoning. Thinking is only offered as a redo option after the initial image, and the user says the difference in results is noticeable even without checking.
The Mac app doesn't have this issue. On web, each attempt wastes a usage credit on an inferior image that will be redone. The poster is asking for workarounds.
More from Models
- DeepSeek V4.1 Leak: 552B New Architecture, HBM Needs Cut 3.9x, Huge Benchmark Gains — teortaxesTex · 2026-09-10
- Anthropic's answer on Claude and the Borisov-Gabber-Vasiu paper sidesteps the key question — giffmana · 2026-09-10
- Google releases TimesFM-3, a 330M multivariate zero-shot forecasting foundation model — AxSaucedo · 2026-09-10
- Leaked DeepSeek V4.1 benchmarks show 552B new-architecture model hitting 63.9 HLE with tools — teortaxesTex · 2026-09-10
- Two months after release, open-weights Kimi K3 shows no documented harm despite cyber capabilities — xeophon · 2026-09-10
- DeepSeek-V4.1-Flash spotted on Hugging Face, unannounced — t4a8945 · 2026-09-10