Shotcraft
Video generation

Pyramid Flow (miniFLUX / SD3)

An autoregressive video generation method based on pyramidal flow matching, trained entirely on open-source datasets and accepted to ICLR 2025. The larger checkpoint produces 10-second video at 768p and 24 fps; a 384p checkpoint produces 5 seconds.

A video generation model by Pyramid Flow (academic collaboration) · read 2026-08-14 · sourcing high

Up to
10 seconds
Resolution
768p
Weights
Downloadable
API
No
Commercial use
Vendor states commercial use is permitted

It also generates 1024p stills.

Commercial use — Vendor states commercial use is permitted

MIT license - the most permissive license of any model in this dataset.

The licence notes recorded during research were read for a commercial-use position and did not state one. Read the licence before you rely on any of this. Nothing on this page is legal advice, and this site does not tell you a model is cleared for commercial use.

Specification

Vendor figures, read 2026-08-14Read from Pyramid Flow (academic collaboration)'s own documentation on 2026-08-14. Vendors change these between model versions without announcing it, so check theirs before you commit a shot to a number. Sourcing for this entry is strong: the figures came from pages we retrieved and read. Check the vendor's current documentation

What it takes in
text, image
What it makes
video, image
Maximum duration
10 seconds
Maximum resolution
768p
Aspect ratios
not established
Controls
text prompt, reference image, resolution checkpoint selection (384p/768p), guidance scale, video guidance scale, CPU offload, sequential offload
Weights
Downloadable
API
No
Pricing
Free open weights (MIT), self-hosting only with no first-party API. Notably light: under 8GB GPU memory with CPU offloading, or under 12GB with standard offloading, per the repository. Training used A100s.
Released or updated
2024-11-13

Supported aspect ratios is not established here. That means no page we could retrieve stated it, not that the model lacks one.

Strengths

  • MIT license, cleanest legal position in open video
  • Trained only on open datasets, so provenance is auditable
  • 10 seconds at 768p/24fps is longer than most 2024-era peers
  • Runs under 8GB VRAM with offloading
  • Peer-reviewed (ICLR 2025)

Weaknesses

  • Visual quality well behind 2025-2026 models
  • Parameter count not stated in the repository
  • Small ecosystem; few LoRAs or downstream tooling
  • Little apparent active development since late 2024

Best for

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

Sources

Read on 2026-08-14. A source that has changed since then will not be reflected here until the next pass.