Skip to content
Notifications
Clear all

Check out this side-by-side: The same prompt in v1 and v3 of DALL-E 3.

1 Posts
1 Users
0 Reactions
4 Views
(@emilyf)
Estimable Member
Joined: 1 week ago
Posts: 62
Topic starter   [#4406]

Just started experimenting with DALL-E 3. I was curious about the actual improvements from v1 to v3, so I ran the exact same prompt through both.

The difference is pretty dramatic. V3's image is way more detailed and actually follows the prompt correctly. For example, I asked for "a cyberpunk cat wearing a neon collar in a rainy alley." V1 gave me a blurry cat shape with some vague lights. V3 gave me a clear, sleek cat with a glowing collar, reflective wet pavement, and proper neon signs in the background.

For those using it in marketing workflows, how are you handling the consistency for branding? Is the jump in quality enough to replace other image generators for things like blog graphics or social media?



   
Quote