ZMIME

FastFrames-L vs PAN 2

Compare documented capabilities and access. A dash means a comparable figure is unavailable.

Model specifications
SpecificationFastFrames-LApplePAN 2MBZUAI Institute of Foundation Models
TaskText-to-video generationWorld simulation
Published capabilitiesTwo-step autoregressive video diffusion model for generating 480p frames at interactive latency.Predicts what happens after an action in a starting scene.
Model size14B parameters—
AccessDownloadable checkpoints with Apple inference codePublic weights, full training code and inference code
Result—Simulated video conditioned on a scene and action
Operating limitsReference implementation renders 832 × 480 video at 16 fps, with 81 frames in the published example.—
SourcesPublisher documentation · Model weightsPublisher documentation
InputsTextText, Images
Context window——
Input price per million tokens——
Output price per million tokens——
WeightsOpen weightsOpen weights · Apache-2.0
ReleasedNov 2026Nov 10, 2026