Independent, side‑by‑side benchmarks for text‑to‑video and image‑to‑video models. Built by filmmakers to help you pick the right model for cinematic shots.