ZMIME

Shot Reader 8B vs TML-d24-varloop-research

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

Model specifications
SpecificationShot Reader 8BVoyagerTML-d24-varloop-researchThinking Machines
TaskShot analysisVariable-depth reasoning research
Published capabilitiesExtracts camera framing, lens, movement and subject positions from footage using a fine-tuned Qwen3-VL model.Researchers can vary the number of middle-layer passes with a single checkpoint.
Model size8B backbone26.4B total / 3.4B active parameters
AccessDownloadable weightsDownloadable research checkpoint
ResultStructured JSON shot sheet—
Operating limits—Native mixed-precision checkpoint occupies approximately 20 GB.
SourcesModel weightsModel weights · Additional documentation
InputsText, Images, VideoText, Images
WeightsOpen weights · Apache-2.0Open weights · Apache-2.0
ReleasedNov 19, 2026Nov 20, 2026