NodeValid

Published video-model weights, licences and hardware notes

Which repositories publish more than one memory figure

Eight of the 15 families publish more than one memory figure. What varies between the figures is different in each case: the offload setting, the frame size, the model size, the card count, the numeric format or the command. As of 2026-09-12.

What varies between two published figuresA pair of figures is informative only when a reader can tell which single thing changed between them, and the register holds six different variables.What changed between the two figuresOne settingReadable as a gradientOffload on and off, or twoframe sizes on one card, sothe difference isattributable.The modelReadable as a cost ofsizeTwo variants at the sameoutput size, which isolateswhat the parameter countadds.Several things at onceNot a comparisonTwo figures from differenttables describe two runsrather than a difference.One variable at a time, or none
Fig. 1 Six variables, each isolated by one vendor, and none of them by all of them.
Families publishing more than one memory figure, by what varies between them. Recorded 2026-09-12.
ItemWhat is publishedDetail
The offload settingOne family9.3G with cpu_offload against 27.5G without
The frame sizeOne family60GB at 720px1280px against 45GB at 544px960px
The model sizeTwo familiesAbout 14.7GB on a 1.3B against around 51.2GB on a 14B, both at 540P; and a three-step ladder from 4GB to 10GB
The card countOne family60.3GB on one GPU against 44.3GB across four or more
The numeric formatOne familyAt least 24GB against at least 12GB on an fp8 build
The commandOne familyA consumer 4090 on one model against at least 80GB on others
The configurationOne family78.55 GB, 77.64 GB and 72.48 GB across three settings
One figure onlyThree familiesA single requirement or floor, with no second point
No memory figure at allFour familiesA timing on a named card, or nothing stated in this column

Inclusion rule. One row per variable a vendor isolated. Families publishing a single figure keep a row, because a single point cannot be read as a gradient. Order. From the variable a reader controls most easily to the least.

1A pair is worth far more than a point

Two figures with one variable between them tell a reader which direction a number moves and by roughly how much. A single figure tells them one configuration and leaves every adjustment to guesswork.

The most useful pairs here hold the output constant and change the model, or hold the card constant and change the frame. Both forms let a reader extrapolate with some confidence.

2Not every pair is a comparison

Where two figures come from different tables at different resolutions on different model sizes, they describe two runs rather than a difference. Reading a ratio out of such a pair produces a number with no meaning.

The register keeps the configuration attached to each figure so that a reader can see which pairs are controlled. Several are, and they are the rows worth planning from.

3The card-count pair is the one most often assumed

One family publishes a single-card figure and a four-card figure for the same output, and the second is about a quarter lower rather than a quarter of the first. A reader assuming linear scaling would be short by a large margin.

Nobody else in the register publishes that pair, so it is the only measured answer here to what sharding buys. It is recorded as one vendor's measurement rather than as a rule.

  • Hardware
    9.3G BF16 with cpu_offload, against 27.5G if CPU offload is not enabledsingle GPU memory usage on the same runrhymes-ai/Allegro on Hugging Face / recorded 2026-09-22
  • Hardware
    Peak GPU memory of 60GB at 720px1280px and 129 frames, and 45GB at 544px960pxthe vendor's own figuresTencent-Hunyuan, HunyuanVideo repository / recorded 2026-09-12
  • Hardware
    About 14.7GB peak VRAM for a 540P video on the 1.3B model, and around 51.2GB on the 14Bthe same output size on two model sizesSkyworkAI, SkyReels-V2 repository / recorded 2026-09-22
  • Hardware
    60.3GB peak GPU memory at 768px on one GPU, and 44.3GB across four or moremeasured on H100 or H800 cardshpcaitech, Open-Sora repository / recorded 2026-09-22
  • Hardware
    For 4.5B models, any machine with at least 24GB of GPU memory is sufficient, and a distilled fp8 configuration works on GPUs with at least 12GBa floor for the model and a lower one for a buildSandAI-org, MAGI-1 repository / recorded 2026-09-22

4Sources

Repository statements and hosted rates are linked from the model pages and the how read page, checked 2026-09-12. Hardware figures are set out on requirements. Nearby: Which variant, Consumer cards, Stated settings.