Pilot datasetOne prompt · one trial · 33 routes · not the official Core rankingRead the limits

openai · O+C

GPT-5.5 Low

gpt-5.5-low

Pilot visual tierA

Human assessment

Distinct stretched-form transition and clean typographic landing.

One diagnostic trial. This is not a durable model capability claim.