Two images & design tools, one decision. Here is the honest split.
| DALL-E | Stable Diffusion | |
|---|---|---|
| Pricing | Freemium | Free plan |
| Best at | Precise compositions, images with readable text, and iterating in plain conversation. | Tinkerers who want control - custom models, local runs, and no per-image costs. |
| In short | OpenAI's image generator, available inside ChatGPT. Strong at following literal instructions and rendering text. | The open-weight image model you can run yourself or through many free interfaces. Infinitely customizable. |
| On Atlas | 3 starter prompts | 3 starter prompts |
Pick DALL-E when the job is Precise compositions, images with readable text, and iterating in plain conversation.. Pick Stable Diffusion when the job is Tinkerers who want control - custom models, local runs, and no per-image costs.. Still torn? Run one real task through both free tiers. An hour of testing beats a week of reviews.
A [subject] in [setting], lit by [light source], shot on a [lens/angle], in a [style] palette of [colors]. No text, no watermark.