Skip to benchmark contentMotionBenchHosted by Baz Studio Human assessment
Pilot datasetOne prompt · one trial · 33 routes · not the official Core rankingRead the limits
anthropic · O+C
Claude Opus 4 6
claude-opus-4-6Pilot visual tierA
Strong semantic compression into a precise wordmark with controlled accent.
One diagnostic trial. This is not a durable model capability claim.