Moonvalley Marey (Realism v1.5)
Marey is Moonvalley's foundational video model, marketed as the first commercially safe video model because it is trained only on licensed high-resolution footage with no scraped content or user submissions.
- Resolution
- 1920x1080 (per the fal.ai Marey endpoint's dimensions enum; Moonvalley publishes no first-party spec table)
- Aspect ratios
- 1920x1080, 1080x1920, 1152x1152, 1536x1152
- Weights
- Not released
- API
- Yes
- Commercial use
- Vendor states commercial use is permitted
Its control surface is aimed at directors: camera control that turns a single image into a navigable 3D scene, motion transfer from a reference clip, trajectory control by drawing an object's path, and keyframing across multiple reference images on a timeline. Moonvalley's own public API is waitlist-only, so programmatic access currently runs through inference partners; on fal.ai the model defaults to 1920x1080 and a 5 second duration.
Moonvalley states Marey is trained only on licensed, high-resolution footage with 'no scraped content, no user submissions, no legal gray zones' and describes outputs as commercially safe with no IP risk.
Specification
- What it takes in
- text, image, video
- What it makes
- video
- Maximum duration
- not established
- Maximum resolution
- 1920x1080 (per the fal.ai Marey endpoint's dimensions enum; Moonvalley publishes no first-party spec table)
- Aspect ratios
- 1920x1080, 1080x1920, 1152x1152, 1536x1152, 1152x1536
- Controls
- image-to-video, camera-control, keyframes, reference-character-consistency
- Weights
- Not released
- API
- Yes
- Pricing
- Not verified. Moonvalley's pricing page renders plan content client-side and exposed no figures when fetched; no first-party per-second or per-clip rate could be confirmed.
- Released or updated
- Marey Realism v1.5 is the variant published on fal.ai as of 2026-08-14; Moonvalley publishes no dated release notes
Maximum duration is not established here. That means no page we could retrieve stated it, not that the model lacks one.
Strengths
- Licensed-only training data with an explicit commercial-safety claim, which is rare
- Trajectory control (draw an object's path) and motion transfer are unusual controls
- Camera control derived from a single image via 3D scene reconstruction
- Positioned and priced for film and advertising post-production workflows
Weaknesses
- First-party public API is waitlist-only, so production access depends on third-party aggregators such as fal.ai
- No first-party spec sheet: duration options, frame rate and native audio support are undocumented
- No published pricing
- Duration defaults to 5s on fal; the full enum of allowed durations was not visible
Best for
- Studios and agencies that need provenance guarantees and director-grade controls, and can live with aggregator-mediated API access.
What to understand before using it
Explainers naming the mechanisms this model's own documentation mentions.
See it from the vendor
Shotcraft hosts no sample output. Judging a model means watching its own showreel with your own eyes, and a clip re-hosted here would be a clip we chose.
Nearby in the directory
Adobe Firefly Video Model
The Adobe Firefly Video Model generates video from a text prompt, optionally guided by an image used as the first or final frame. Adobe's Firefly Serv
Video generation
Allegro
A small text-to-video model comprising a 2.8B VideoDiT and a 175M VideoVAE, released under Apache-2.0 in October 2024. It generates 88-frame clips at
Video generation
Amazon Nova Reel 1.1
Amazon Nova Reel generates video from text prompts or from text plus an image, and is available only through Amazon Bedrock. AWS documents output at 1
Video generation