Frame-level temporal control
Meridian 4 was trained on densely captioned footage, which makes imaginative transitions and precise key-framing of individual elements in a scene possible.
Meridian 4 is the first of a new generation of foundation models trained by Orrery on infrastructure built for large-scale multimodal training. It is a marked improvement in fidelity, consistency, and motion over Meridian 2, and a step toward general world models.
Every clip on this page was generated with Meridian 4, with no post-processing.
orrery
Trained jointly on video and stills, Meridian 4 will power Orrery’s Text-to-Video, Image-to-Video, and Text-to-Image tools, existing control modes such as Motion Paint, Advanced Camera Rigs, and Sequence Mode, as well as upcoming tools for finer control over structure, style, and motion.
Meridian 4 ships with a new set of safeguards, including our rebuilt in-house visual moderation system and C2PA provenance standards.
orrery
orrery
orrery
orrery
orrery
orrery
orrery
orrery
orrery
orreryMeridian 4 was trained on densely captioned footage, which makes imaginative transitions and precise key-framing of individual elements in a scene possible.
Meridian 4 excels at expressive human characters with a wide range of actions, gestures, and emotions, which opens up new storytelling options.
Meridian 4 is a collaboration between our research team, engineers, and a resident group of directors. It was shaped to interpret a broad range of styles and cinematic language.
orrery
orrery
orrery
orrery
orrery
orrery
orrery
orreryAs part of the Meridian family, we have been working with production companies and media organizations to build custom versions of Meridian 4. Customization allows for more stylistically controlled and consistent characters, and targets the specific artistic and narrative requirements of a project, among other features. If your company is interested in fine-tuning and custom models, reach out using the form below:


orrery