NodeValid

Published video-model weights, licences and hardware notes

MAGI-1: eight H100s or one 4090, by size

The repository states H100 or H800 x 8 for the 24B model and RTX 4090 x 1 for the 4.5B. It adds that the 4.5B needs at least 24GB, and that a distilled fp8 configuration works with at least 12GB. As of 2026-09-22.

MAGI-1 in the hardware column, as publishedStating hardware as a count of a named card covers memory, numeric formats and interconnect in one line, which a gigabyte figure cannot do.The memory statement as publishedH100 or H800 x 8 for the24BOne RTX 4090 for the 4.5B; 12GB for a distilled fp8 build720 x 720 as the 4.5BdefaultSet for speed; arbitrary resolutions described as supportedNo ceiling recordedFrames are an argument with no stated maximumA card count carries the interconnect too
Fig. 1 Cards counted rather than gigabytes quoted, with a separate floor for the fp8 build.
What the repositories actually stateFilled where the repository states something, hollow where it does not. A blank is not an inference.What the repositories actually stateWhat the pages stateWhat the pagesstateHardware, in the vendor's own wordsHardware, in the vendor's own words — What the pages state: H100 or H800 x 8 for the 24BThe detail published beside itThe detail published beside it — What the pages state: One RTX 4090 for the 4.5B; 12GB for a distilled fp8 buildResolution, from the same pagesResolution, from the same pages — What the pages state: 720 x 720 as the 4.5B defaultLength, from the same pagesLength, from the same pages — What the pages state: No ceiling recorded
Fig. 2 Filled where the repository states something, hollow where it does not. A blank is not an inference.
MAGI-1 in the hardware column, with the columns that qualify it. Read from github.com/SandAI-org on 2026-09-22.
ColumnWhat the pages state
Hardware, in the vendor's own wordsH100 or H800 x 8 for the 24B
The detail published beside itOne RTX 4090 for the 4.5B; 12GB for a distilled fp8 build
Resolution, from the same pages720 x 720 as the 4.5B default
Length, from the same pagesNo ceiling recorded

Inclusion rule. The value this page is about, the detail the vendor attached to it, and the two columns that change how it should be read. A column the vendor left empty keeps its row and says so. Order. This column first, then its detail, then the two columns that qualify it.

1Counting cards says more than counting gigabytes

Eight H100s is not just 640GB of memory. It is a node with a fast interconnect, a driver stack and formats the run was exercised against. Written as a memory total it would be unreproducible.

One RTX 4090 works the same way in the other direction: consumer memory, consumer bandwidth, and a format set that is not the same as a data-centre card's.

2Three floors for two sets of weights

At least 24GB for the 4.5B model itself, at least 12GB for a distilled fp8 build of it, and eight data-centre cards for the 24B. The middle figure is about software, not weights.

That is the clearest statement in the register that a memory requirement describes a build. The parameter count does not change between 24GB and 12GB; the numeric format does.

3What the resolution default tells you

The 4.5B default is 720x720, described as set to accelerate inference, with arbitrary resolutions supported. So the hardware figures belong to a square default rather than to the output a reader may want.

Length is unbounded in the repository: frames are an argument with no stated maximum. Neither of the qualifying columns pins these figures down as tightly as the card counts suggest.

  • 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
  • Hardware
    H100/H800 x 8 for the 24B model, RTX 4090 x 1 for the 4.5Bstated as a count of a named cardSandAI-org, MAGI-1 repository / recorded 2026-09-22

4Sources

Every figure above is the vendor's own, taken from github.com/SandAI-org on 2026-09-22 with nothing measured here. The rest of the release is on its family entry, the whole column is on hardware, and all memory figures together are on requirements. Nearby: 32.54 GB, three architectures, No card, no figure.