How many frames, views and what resolution does each text/image/video-to-4D system produce, and what GPU time and memory do the authors report?
Corrected 2026-10-23 - see C3-B588-07; the original statements below are kept for history and are superseded where the correction says so. Partially verified: the curator re-read none of the cited sources; S1, S11, S3, S4, S13, S14, S15, S16, S17, S5, S6, S8 rest on the producer's pinned reading. Reported per-asset cost spans about four to five orders of magnitude, all author-measured on different hardware. Seconds: L4GM (0.3 s per 16-frame pass plus about 2 s for views, RTX 4080 Super). Minutes: DreamGaussian4D (6.5-10 min, 80 GB A100), Diffusion4D (8 min, paper), SV4D (15-20 min fitting) and SV4D 2.0 (1 min generation + 21 min fitting, up to 30 GB, H100). Hours: Consistent4D (about 2-2.5 h, V100), MAV3D (about 6.5 h on 8 A100) and 4D-fy (about 23 h on one 80 GB A100). Clips are short: 12-32 frames per pass depending on the system (12, 14, 16, 21, 24, 32), time axes are continuous for the NeRF/neural-field systems, and no passage read states output duration in seconds for text-to-4...
Structured fields
{
"citations": [
{
"url": "https://arxiv.org/html/2301.11280v1",
"note": "candidate evidence excerpts"
},
{
"url": "https://arxiv.org/html/2312.17142v3",
"note": "candidate evidence excerpts"
},
{
"url": "https://arxiv.org/html/2311.17984v2",
"note": "candidate evidence excerpts"
},
{
"url": "https://github.com/sherwinbahmani/4dfy",
"note": "candidate evidence excerpts"
},
{
"url": "https://arxiv.org/html/2311.02848v1",
"note": "candidate evidence excerpts"
},
{
"url": "https://github.com/yanqinjiang/consistent4d",
"note": "cited at https://github.com/yanqinJiang/Consistent4D; candidate evidence excerpts"
},
{
"url": "https://arxiv.org/html/2407.17470v2",
"note": "candidate evidence excerpts"
},
{
"url": "https://arxiv.org/html/2503.16396v3",
"note": "candidate evidence excerpts"
},
{
"url": "https://github.com/stability-ai/generative-models",
"note": "cited at https://github.com/Stability-AI/generative-models; candidate evidence excerpts"
},
{
"url": "https://arxiv.org/html/2406.10324v1",
"note": "candidate evidence excerpts"
},
{
"url": "https://github.com/nv-tlabs/L4GM-official",
"note": "candidate evidence excerpts"
},
{
"url": "https://arxiv.org/html/2405.16645v1",
"note": "candidate evidence excerpts"
},
{
"url": "https://github.com/Stability-AI/generative-models",
"note": "observed URL (dated/alias form of the cited catalogue source; canonical identity retained)"
},
{
"url": "https://github.com/yanqinJiang/Consistent4D",
"note": "observed URL (dated/alias form of the cited catalogue source; canonical identity retained)"
}
]
}Evidence claims
- [sourced_fact] Reported per-asset cost spans about four to five orders of magnitude, all author-measured on different hardware. Seconds: L4GM (0.3 s per 16-frame pass plus about 2 s for views, RTX 4080 Super). Minutes: DreamGaussian4D (6.5-10 min, 80 GB A100), Diffusion4D (8 min, paper), SV4D (15-20 min fitting) and SV4D 2.0 (1 min generation + 21 min fitting, up to 30 GB, H100). Hours: Consistent4D (about 2-2.5 h, V100), MAV3D (about 6.5 h on 8 A100) and 4D-fy (about 23 h on one 80 GB A100). Clips are short: 12-32 frames per pass depending on the system (12, 14, 16, 21, 24, 32), time axes are continuous for the NeRF/neural-field systems, and no passage read states output duration in seconds for text-to-4D. View coverage is object-centric: SV4D/2.0 use elevation and azimuth cameras and fail on strong perspective; L4GM assumes 0 degree input elevation. · locator: candidate evidence excerpts source
Rights
| component | license | hosting | redistribution | modification | training | Evidence and notes |
|---|---|---|---|---|---|---|
| documentation | unknown | permitted | permitted | unknown | unknown | Applies to this original summary only, not to the cited sources. |
Linked records
- cites → https://arxiv.org/html/2301.11280v1
- cites → https://arxiv.org/html/2311.02848v1
- cites → https://arxiv.org/html/2311.17984v2
- cites → https://arxiv.org/html/2312.17142v3
- cites → https://arxiv.org/html/2405.16645v1
- cites → https://arxiv.org/html/2406.10324v1
- cites → https://arxiv.org/html/2407.17470v2
- cites → https://arxiv.org/html/2503.16396v3
- cites → https://github.com/nv-tlabs/l4gm-official
- cites → https://github.com/sherwinbahmani/4dfy
- cites → https://github.com/stability-ai/generative-models
- cites → https://github.com/yanqinjiang/consistent4d