What is P-Image-Ideogram?
P-Image-Ideogram is Ideogram's fast text-to-image model, built with Pruna AI. At 1K it typically returns in about 2.9–5.0 seconds for roughly $0.003–$0.015 per image, depending on Quality, with a Very High level available at $0.033 per 1K image for the toughest briefs. It is built for high-volume production and iteration without jumping to slower premium models.
What is Quality (Very Low, Low, Medium, High, Very High)?
Quality is a five-level control that trades off speed, price, and how much compute goes into the prompt. Levels are not a strict worse-to-better ladder: match Quality to how complex or precise the brief is. Medium is the everyday default; use High for dense typography, complex prompts, and fine detail; use Very High when you want maximum compute on the hardest briefs; use Very Low or Low for drafts and A/B exploration.
How does P-Image-Ideogram compare on speed and price?
On human-preference and automated judge benches, quality lands near leading image models while staying much faster and cheaper than GPT-class and Nano Banana Pro runs (often about a second and under ~$0.02 at 1K, versus tens of seconds and $0.10+). Figures are directional.
Is there an API for this?
Yes. POST a prompt to /v1/text-to-image/p-image-ideogram with an Api-Key header. Optional fields include quality (VERY_LOW, LOW, MEDIUM, HIGH, VERY_HIGH), resolution (1K or 2K), and aspect_ratio. Defaults are quality=MEDIUM and resolution=1K. See developer.ideogram.ai for documentation.