DALL-E or FutureSelf for vision board images?
Inside ChatGPT, DALL-E 3 was replaced in March 2025 by GPT-4o image generation, which is available on the free plan and can edit a photo you upload. FutureSelf does one job: one selfie becomes 8 photoreal scenes of you on one board, $4.90 once, first scene free, no prompting.
Flexible chat or a finished board
ChatGPT can make any image you can describe, and its free plan is enough to try. FutureSelf is narrower: it turns one selfie into a finished 8-scene board without prompting. If you enjoy iterating in a chat, ChatGPT is the more flexible tool.
Why DALL-E 3 isn't enough
Eight prompts, eight results
A full board in a chat means describing yourself and the style again for every scene, and checking each one.
Layout is on you
You get separate images. Arranging them into a board and sizing it for a wallpaper is a separate job.
What changed
In March 2025 ChatGPT moved from DALL-E 3 to image generation built into GPT-4o, and free users got access with daily limits. It can also edit a photo you upload, which DALL-E 3 in ChatGPT could not. Checked 2026-10-12.
One selfie. 8 photoreal scenes of you on one board.
Where FutureSelf is better
Effort and a finished result: one selfie, a few answers, 8 scenes on one board sized for a phone or desktop, drawn by Google's Gemini image model.
One selfie. 8 photoreal scenes of you on one board.
Common Questions
Q.Does FutureSelf use DALL-E?
No. FutureSelf's scenes are drawn by Google's Gemini image model. We do not train or fine-tune a model.
Q.Can I use ChatGPT to write my description for FutureSelf?
You can, though you do not need to: the questions in FutureSelf are enough.
See one scene of you, free
Make my vision boardFirst scene free, no card. A free account is needed.