GPT-Image-2.5 aces complex prompts but falls short of Midjourney's aesthetic sensibility
perilli · x · 2026-09-12
An industry reviewer tested OpenAI's new GPT-Image-2.5: its ability to follow complex prompt instructions is impressive, but its creative and aesthetic sensibility remains far behind Midjourney models—bigger improvements are needed to wow the media and entertainment industry. Even after extensive iteration with GPT-6 Astra Mac and ultimately GPT-6 Astra Ultra, the result still wasn't perfect.
More from Models
- Hands-on with Meta's Muse Agent: generous limits, clunky credential handoff — brandon_galang · 2026-09-13
- Text-only reasoning solves Rubik's Cube: GPT-5.6 Sol clears CubeBench's 0% barrier in 333 moves — Willing_Plate_5417 · 2026-09-12
- Dev demos Google Astra nailing live assistance on Rust/wgpu procedural textures — Dimillian · 2026-09-12
- GPT-6 Astra wild demos: race hand-drawn horses, papers turned into Blender scenes — socialwithaayan · 2026-09-12
- Mystery model Agnes 3.0 Flash tops efficiency index, a 33B dense rival to Qwen — Informal-Trouble2183 · 2026-09-12
- Nate Silver: Ignore AI benchmarks that labs already know in advance — soumitrashukla9 · 2026-09-12