bestaiq
// IMAGE

DALL·E 3

OpenAI's text-to-image model built natively into ChatGPT

5.9/10
Fair · T3
SUB-SCORE SPINE
7.5
CAP
6.0
VAL
9.0
EAS
7.0
PRI
3.5
SUP
3.5
ECO
Independent · ad-free verdicts · we may earn affiliate commissions — this never affects our scores.
TABLE · AT A GLANCE
Category IMAGE [tools] ✓2026-06
Starting price $0/mo [tools] ✓2026-06
Pricing model Freemium [tools] ✓2026-06
Free tier Yes [tools] ✓2026-06
Company OpenAI [tools] ✓2026-06
Founded 2023 [tools] ✓2026-06
Website openai.com/index/dall-e-3 [tools] ✓2026-06
Our score 5.9/10 · Fair · T3 [tools] ✓2026-06

//Strengths & things to watch

STRENGTHS
  • Follows complex, detailed prompts closely
  • Renders in-image text, hands, and faces better than predecessors
  • Conversational iteration via ChatGPT
  • Users own outputs (reprint, sell, merchandise)
THINGS TO WATCH
  • Removed from the OpenAI API on May 12, 2026 (no longer programmatically available)
  • Photorealistic human faces can look synthetic
  • Mandatory GPT-4 prompt rewriting reduces exact-prompt control
  • Heavy content refusals (named public figures, living-artist styles)
FIG · KEY FEATURES

//What users say

Cross-referenced synthesis of 5 external sources · not on bestaiq

MOST PRAISED
  • Closely follows long, detailed prompts
  • Renders legible in-image text, hands, and faces better than predecessors
  • Conversational refinement built natively into ChatGPT
  • Outputs are owned by the user (reprint/sell/merchandise)
MOST CRITICIZED
  • Removed from the OpenAI API on 2026-05-12 (no longer programmatically available)
  • Replacement GPT Image models reported to lose DALL·E 3's painterly/material rendering
  • Photorealistic human faces can look synthetic / rendered
  • Mandatory GPT-4 prompt rewriting reduces control over exact prompts
  • Heavy content refusals (public figures, living-artist styles)
TAKEAWAYLoved for prompt accuracy, text rendering, and ChatGPT convenience; now legacy after API removal.

//Related tools