Vibeleaderboard
← All Intel
Intel / post

Ideogram 4.5 ranks #23 in image editing with tiered per-image pricing

Source
x.com
Date
ArtificialAnlys@ArtificialAnlys

Ideogram 4.5 debuts at #23 on the AA-Image-Editing v2.0 leaderboard, Ideogram's first model on the board, and lands at #34 on AA-Image-T2I v2.0 Ideogram 4.5 is Ideogram's new image model, released on September 30. Ideogram positions it as a high precision edit model, built to reduce pixel shifts, color changes and texture artifacts across multi-turn edits. It also supports Text to Image. We evaluated Ideogram 4.5 at “high” quality level. In AA-Image-Editing v2.0, Ideogram 4.5 takes #23, just behind HiDream-O1-Edit-1.5 and HunyuanImage 3.0 Instruct and ahead of ByteDance's Seedream 5.0 Lite. It is the first Ideogram model on our Image Editing Leaderboard. In AA-Image-T2I v2.0 it ranks #34, climbing from #40 for Ideogram 4.0. At the High quality setting we evaluated, Ideogram 4.5 is priced at $0.22 per image ($220 per 1k images) for edits and $0.10 per image ($100 per 1k) for Text to Image. Lower quality tiers start at $0.008 per image for edits and $0.03 per image for Text to Image. Ideogram 4.5 is live in Ideogram, the Ideogram API and Ideogram's launch partners, with open weights coming soon. Congratulations to @ideogram_ai on the release! See below for our analysis and…

Read the full post on X

Context

Ideogram 4.5's reported image-editing rank belongs to a specific quality setting and price. In its October 3 evaluation report, Artificial Analysis placed the model 23rd on its image-editing leaderboard at the high quality setting it evaluated, priced at $0.22 per edit. The report also listed lower-priced tiers starting at $0.008 per edit, but those were not the setting behind that result. Anyone comparing the model's ranking with another service's price needs to keep the tested configuration attached to the number.

Terms in this piece · Glossary
  • eval — A repeatable test for AI quality — a set of tasks plus scoring — used the way software teams use test suites, because model output is too variable to judge by eyeballing.
  • open weights — A model whose trained parameters are published for anyone to download and run — unlike API-only models you can access but never possess.
More from ArtificialAnlys
Recommended reads
Comments

Checking sign-in…

Loading comments…