#GPT-5 Image Generation Fails Silently — Latency, Unresponsive UI, and No Render Output (Oct 9–15 202

1 messages · Page 1 of 1 (latest)

tame depotBOT
#

Reported by @pine plover

Bug Report: GPT-5 Image Generation Fails Silently — Latency, Unresponsive UI, and No Render Output (Oct 9–15 202
`Steps to Reproduce`

Open ChatGPT (GPT-5 model) on ChatGPT Plus account.

Request any image generation (e.g., generate a photoreal image of a dog in daylight).

Observe: no output, no error, and increasing UI latency.

Attempted on Chrome 129 and Safari 17 on macOS Sequoia 15.6.1 (M2) — identical result.

Retried multiple sessions; same across browsers, devices, and stable fiber (433 / 176 Mbps).

Tracked log file available

`Expected Result`

GPT-5 image generation should execute normally, return a generated image, and display confirmation logs or processing progress without significant lag.

`Actual Result`

No image is generated; response window stalls for 60–120 s with “Page Unresponsive” alerts, connection-lost banners, or delayed text responses.
No error message or completion ID returned.
GPT-4o fallback works intermittently but ignores detailed prompt directives (lighting, realism parameters).

`Environment`

macOS Sequoia 15.6.1 (Apple M2, MacBook Air), 357 GB free of 494 GB Browsers: Chrome 129, Safari 17 Fiber Internet: 433 Mbps down / 176 Mbps up (verified stable) Issue window: Oct 9 – 15 2025 (Pacific Time) ChatGPT Plus (GPT-5 model) Tested daily > 12 times; 100% failure rate; GPT-4o partial fallback confirmed

#
Additional Information

Please provide relevant details to help resolve the issue, such as:

  • ChatGPT Shared Link (if applicable).
  • Screenshots or videos demonstrating the problem.

-# ➜ Need to contact support? Visit the OpenAI Help Center.

pine plover