Skip to benchmark contentMotionBenchHosted by Baz Studio Human assessment
Pilot datasetOne prompt · one trial · 33 routes · not the official Core rankingRead the limits
openrouter · O+C
Deepseek V4 Pro
deepseek-v4-proPilot visual tierC
Large areas read as nearly blank; semantic handoff is weak.
One diagnostic trial. This is not a durable model capability claim.