Short answer
The reviewed OpenAI guidance does not establish a dependable standard time for editing an existing image in ChatGPT. OpenAI says image generation may take a few minutes depending on request complexity, but that is general generation guidance—not an edit-time estimate or maximum wait. [1]
An edit may also involve reviewing the result and asking for a follow-up change. OpenAI describes those editing options, but does not give a duration for one edit or for a multi-step editing task. [1]
What the timing guidance does—and doesn’t—tell you
OpenAI’s ChatGPT Images help article says image generation may take a few minutes depending on request complexity. It does not provide a typical duration, a maximum wait, or a separate estimate for editing an uploaded image. So that guidance should not be converted into an expected time for an edit. [1]
Third-party descriptions such as “fast” are subjective, not timed results. User reports of long waits are individual experiences, mostly about image generation; they do not establish how long an edit should take. Treat them as anecdotes, not a benchmark.
ChatGPT editing versus an image-editing API
In ChatGPT, OpenAI says you can upload an existing image and describe a change, or select an area to edit. The reviewed help guidance does not give a processing-time range for either approach. [1]
For production use, OpenAI’s API documentation describes two options:
- Image API: The image-edit guide describes editing an existing image with a prompt. The API reference specifies
POST /v1/images/editsfor image edits and says the endpoint supports GPT Image models. [2][4] - Responses API: Use the image-generation tool for a conversational or multi-step workflow, including iterative edits. [2]
The image-generation guide names gpt-image-2.5-sunburst and gpt-image-2.5-flare for direct Image API requests. The edit reference confirms GPT Image model support, but the reviewed documentation does not establish that Sunburst is more precise for editing, so this article does not recommend it on that basis. OpenAI’s separate Images 2.6 announcement describes 2.6 API options; it does not, by itself, confirm those options for the documented 2.5 edit path. Check the current API reference for the model IDs applicable to your request. None of these sources supplies a dependable edit-time estimate. [2][3][4]
For an API decision, don’t substitute ChatGPT interface experience or image-generation timings for edit-specific measurements. If no published timing commitment fits your use case, test representative images and prompts. Record request start and completion, model and settings, and whether each request succeeds or needs a retry. Compare repeated runs for your workload; don’t present a small pilot as a universal service guarantee.
Why a complete edit may take longer than one request
The time to finish a task can include more than processing one request: you may review the output and ask for another edit. OpenAI describes iterative editing in the Responses API, but does not quantify the time for each round. [1][2]
For API requests, OpenAI’s guide says higher image complexity can increase latency. That is useful context for testing API workflows, but it does not establish a typical duration for an edit in ChatGPT or a guaranteed API response time. [2]
Bottom line: The reviewed OpenAI guidance does not establish a dependable standard duration for editing an existing image in ChatGPT. Its “few minutes” statement concerns image generation, not an edit estimate or maximum. For production API work, measure representative edit requests rather than relying on generation figures or user anecdotes.
Sources
Images in ChatGPT — OpenAI Help Center