Skip to benchmark contentMotionBenchHosted by Baz Studio Human assessment
Pilot datasetOne prompt · one trial · 33 routes · not the official Core rankingRead the limits
openai · O+C
GPT-5.5 High
gpt-5.5-highPilot visual tierA
Strong beam-to-type morph with confident scale and timing.
One diagnostic trial. This is not a durable model capability claim.