Models · Multi-view depth

VGGT

VGGT (Meta): cameras, depth and 3D points for many images in one pass. Non-commercial weights.

stable Fork of facebookresearch/vggt@a288dd0 · backends: portable, kernels

Checkpoint vggt-1b

Paper 2025 Added in 0.1.0 Evaluated with 0.2.0

Kernels, RTX 5060 Ti 16 GB 460 ms 336x518, fp16
vs the faster original 2.1× original fp16: 965 ms
Peak memory 3,650 MiB original: 8,688 MiB
Accuracy vs original fp32 ±0.0000 kernels AbsRel, DTU scans 1 (0.0026 original)

Use it

pip install fastforks

Lean install with only what this model needs: pip install fastforks-core

from fastforks import multiview

model = multiview.load("vggt-1b")  # kernels on a supported GPU, else portable PyTorch
result = model(images)
result.depth  # also: confidence, world_points, intrinsics, cam_to_world, points_confidence
result.backend  # what ran: mode, device, dtype, kernels

The weights are non-commercial: fastforks downloads them only with FASTFORKS_NONCOMMERCIAL=1 set.

fastforks bench vggt-1b

Times the model on this machine. load(…, mode="portable") or mode="kernels" picks a backend; usage guide.

Speed and memory

Every backend on one GPU (RTX 5060 Ti 16 GB) and the same input, the authors’ code included. Kernels runs fp16. Portable runs fp16 on CUDA.

Input: synthetic, 336x518

BackenddtypeTime per call (ms)vs original fp16Peak memory
allocated / reserved
Load (s)First call (s)
Original fp32float32 (the authors' autocast off)fp32 25020.39×6,958 MiB/7,358 MiB14.22.50
Original fp16original default (README / demo): torch.autocast(bfloat16)fp16 965–8,688 MiB/9,158 MiB19.41.29
Portable fp32fp32not measured ––––
Portablefp16 6511.5×3,643 MiB/3,874 MiB6.81.02
Kernelsfp16 4602.1×3,650 MiB/4,194 MiB7.00.75
Portable + torch.compile 5571.7×3,611 MiB/3,856 MiB74 (build)–
TensorRT fp16fp16 4242.3×3,642 MiBenginenot comparable199 (build)–

Measured on RTX 5060 Ti 16 GB, torch 2.14.1+cu130, 2026-11-03 to 2026-11-16. First call, 3 warm-up, peak reset, median of 3 rounds x median of 5. Ratios divide two times of this table (same GPU, same input). Original fp16 is the authors’ own precision option, else torch.autocast. torch.compile and TensorRT: build time instead of load time. TensorRT memory (engine, not comparable) is the GPU memory the engine took, read from the driver because TensorRT allocates outside PyTorch, so it can't be compared with the PyTorch peaks of the other rows.

Other devices

DeviceBackenddtypems / callPeak memoryOutput vs referenceInput
Apple M1 Pro 16 GB MPS, torch 2.14.1, 2026-11-14Originalfp327987 10.1 GiB processmatches336x518
Portablefp166638 8,423 MiB processmatches336x518
Portablefp327209 10.4 GiB processmatches336x518
Kernelsfp327016 10.4 GiB processmatches336x518
NVIDIA L4 CUDA, torch 2.14.1+cu130, 2026-11-07Portablefp16663 3,619 MiB/3,834 MiB–336x518
Kernelsfp16660 3,640 MiB/4,134 MiB–336x518
NVIDIA L40S CUDA, torch 2.14.1+cu130, 2026-11-07Portablefp16196 3,619 MiB/3,834 MiB–336x518
Kernelsfp16189 3,640 MiB/4,134 MiB–336x518
NVIDIA A10 CUDA, torch 2.14.1+cu130, 2026-11-07Portablefp16517 3,619 MiB/3,834 MiB–336x518
Kernelsfp16504 3,630 MiB/4,174 MiB–336x518
NVIDIA A100-SXM4-40GB CUDA, torch 2.14.1+cu130, 2026-11-08Portablefp16190 3,619 MiB/3,834 MiB–336x518
Kernelsfp16184 3,634 MiB/4,134 MiB–336x518
NVIDIA RTX PRO 6000 CUDA, torch 2.14.1+cu130, 2026-11-08Portablefp16104 3,643 MiB/3,874 MiB–336x518
Kernelsfp16106 3,654 MiB/4,194 MiB–336x518

CUDA memory is torch’s peak allocated / reserved; Apple has no such counter, so the Mac rows show the process footprint (weights, Python and torch included). “Output vs reference” compares the device’s output with the stored golden output under the card’s tolerance; – = not checked on that device. Kernels “not implemented” = no kernels for this model on that backend yet: it runs portable there.

Licence and weights

Non-commercial only

Code
LicenseRef-VGGT LICENSE
Weights
CC-BY-NC-4.0 source
Training data
Per the paper (arXiv 2503.11651): Co3Dv2, BlendMVS, DL3DV, MegaDepth, Kubric, WildRGB, ScanNet, HyperSim, Mapillary, Habitat, Replica, MVS-Synth, PointOdyssey, Virtual KITTI, Aria Synthetic Environments, Aria Digital Twin; DINOv2 initialisation
  • weights: CC-BY-NC-4.0
  • The VGGT code licence allows commercial use, but the VGGT-1B weights are CC-BY-NC-4.0; Meta's separately licensed commercial checkpoint is not packaged.

Our reading of the licences, not legal advice.

Weights

model.pt 5.0 GB· hosted on Hugging Face

fastforks downloads and checks these files on first use (fastforks download vggt-1b). Non-commercial weights need FASTFORKS_NONCOMMERCIAL=1.

Accuracy

The test set through the authors’ code and through ours: same checkpoint, same protocol. The distance between the original’s fp32 and fp16 runs is what half precision alone costs; our runs should stay within it.

DTU scans 1

9, 24, 33, 75, view 24 + 7 sources, AbsRel / delta1 %

RundtypeAbsRel ↓δ1.25 % ↑Δ AbsRel vs fp32 originalscansms / scan
Original fp32fp320.002699.94 reference 5 3422
Original fp16bf160.002799.94 −0.0001 5 1253
Portable fp32fp320.002699.94 ±0.0000 5 3400
Portablefp160.002699.94 ±0.0000 5 856
Kernelsfp160.002699.94 ±0.0000 5 604

640x512 crops, the 7 most co-visible views as sources; GT = depth rendered from the structured-light points. AbsRel / delta1 over GT pixels, after per-image median scaling unless the model takes the cameras (calibrated MVS: cameras, error in mm; PatchMatchNet also the depth range 425-935 mm, MVSAnywhere its own range from the camera baselines). Original = the authors’ code, same checkpoint and protocol. Measured on original fp32, original fp16, portable, kernels: RTX 5060 Ti, torch 2.14.1+cu130, 2026-11-06; portable fp32: RTX 5060 Ti, torch 2.14.1+cu130, fastforks 30c7a8e9, 2026-11-17.

Closeness to the original

Beyond the test-set scores: the raw output of ours against the original’s fp32 output on the same input, pixel by pixel. CPU, float32: max |diff| <= 0.002 vs the original. GPU, float16 (the default), portable and kernels: mean |diff| <= 0.02 and p99 <= 0.25 (absolute error of the output depth).

CheckToleranceMeasured
CPU fp32, max |diff|≤ 0.002 not recorded
GPU default dtypes, mean |diff|≤ 0.02 0.00044
GPU default dtypes, p99 |diff|≤ 0.25 0.00183

Recorded for vggt-1b on RTX 5060 Ti, torch 2.14.1+cu130, 2026-10-31, with these kernels: cudnn_haccfastbb

Details

VGGT, Visual Geometry Grounded Transformer (Wang et al. 2025): cameras, depth maps and point maps of every view from unposed images in one forward pass. A DINOv2 ViT-L patch embedding feeds 24 alternating frame / global attention blocks; a camera head and two DPT heads decode the tokens.

Input / output
Input: several images of one scene (posed and calibrated for multi-view stereo, unposed otherwise). Output: depth maps and, where the model predicts them, cameras and 3-D points.
Input size
any H x W >= 1, resized or padded inside the model; tested up to 3000x4000 (8,219 MiB allocated)
Options
  • long_side=518>= 14· changes the output
Outputs
depth, confidence, world_points, intrinsics, cam_to_world, points_confidence(on request)
CUDA kernels
  • fastbbfast backbone kernels: flash attention, window attention, instance norm and rotary embeddings for ViT and similar backbones
  • cudnn_haccconvolutions with fp16 accumulation (cuDNN)
Apple (MPS)
vggt_metalMetal: VGGT rotary embeddings. fp32, fp16
Full reference, generated from the repository

VGGT (Meta): cameras, depth and 3D points for many images in one pass. Non-commercial weights.

Commercial use: no (weights: CC-BY-NC-4.0).

variantinputbackendtime (ms)peak memory (MiB, allocated / reserved)
vggt-1bsynthetic (336x518)original (fp32)
original
portable
kernels
portable + torch.compile
TensorRT fp16
2502
965
651
460
557
424
6958 / 7358
8688 / 9158
3643 / 3874
3650 / 4194
3611 / 3856
6421 / 6472
Details
  • Original code: https://github.com/facebookresearch/vggt at a288dd0
  • Code licence: LicenseRef-VGGT (permissive); files: LICENSE
  • Weights licence: CC-BY-NC-4.0
  • Licence evidence: https://github.com/facebookresearch/vggt/blob/a288dd0/LICENSE.txt and https://huggingface.co/facebook/VGGT-1B (license: cc-by-nc-4.0)
  • Training data: Per the paper (arXiv 2503.11651): Co3Dv2, BlendMVS, DL3DV, MegaDepth, Kubric, WildRGB, ScanNet, HyperSim, Mapillary, Habitat, Replica, MVS-Synth, PointOdyssey, Virtual KITTI, Aria Synthetic Environments, Aria Digital Twin; DINOv2 initialisation
  • Licence note: The VGGT code licence allows commercial use, but the VGGT-1B weights are CC-BY-NC-4.0; Meta's separately licensed commercial checkpoint is not packaged.
  • Options: long_side = 518 (>= 14)
  • Kernel libraries: fastbb, cudnn_hacc
  • Variants: vggt-1b
  • Outputs: depth, confidence, world_points, intrinsics, cam_to_world, points_confidence (in extras); kind: scale_invariant
  • Input size: any H x W >= 1, resized or padded inside the model
  • Tested up to (portable, fp16, RTX 5060 Ti 16 GB): vggt-1b 3000x4000 (8.0 GiB peak allocated, 10.1 reserved)
  • vggt-1b 336x518, first call: portable + torch.compile 74 s (cold Inductor cache), TensorRT fp16 199 s (ONNX + engine).
  • vggt-1b original (fp32) 336x518: float32 (the authors' autocast off)
  • vggt-1b original 336x518: original default (README / demo): torch.autocast(bfloat16)
  • vggt-1b load / first call: original 19.37 s / 1.289 s (336x518), portable 6.78 s / 1.025 s (336x518), kernels 7.03 s / 0.751 s (336x518)
  • Protocol, median of 5 after 3 warm-up, uint8 CUDA input (synthetic for bench_inputs sizes), card defaults; gc before each load; peak = max over the first call and the timed calls

Measured

640x512 crops, the 7 most co-visible views as sources; GT = depth rendered from the structured-light points. AbsRel / delta1 over GT pixels, after per-image median scaling unless the model takes the cameras (calibrated MVS: cameras, error in mm; PatchMatchNet also the depth range 425-935 mm, MVSAnywhere its own range from the camera baselines). Original = the original code named under Details, same checkpoint and protocol.

variantdatasetoriginal fp32original fp16portableportable fp32kernelsnote
vggt-1bDTU scans 1, 9, 24, 33, 75, view 24 + 7 sources, AbsRel / delta1 %0.0026 / 99.940.0027 / 99.940.0026 / 99.940.0026 / 99.940.0026 / 99.94

Measured on portable: NVIDIA GeForce RTX 5060 Ti, torch 2.14.1+cu130, 2026-11-06; portable32: NVIDIA GeForce RTX 5060 Ti, torch 2.14.1+cu130, fastforks 30c7a8e9, 2026-11-17; kernels: NVIDIA GeForce RTX 5060 Ti, torch 2.14.1+cu130, 2026-11-06.

Parity

CPU, float32: max |diff| <= 0.002 vs the original. GPU, float16 (the default), portable and kernels: mean |diff| <= 0.02 and p99 <= 0.25 (absolute error of the output depth).

variantkernels error vs original, mean / p99GPU, torchkernels used
vggt-1b0.00044 / 0.00183RTX 5060 Ti, torch 2.14.1+cu130cudnn_hacc@cudnn92400, fastbb@7d058322634d:sm_120

Example output

One fixed input through this checkpoint, the same for every model of the task (kernels · float16 · NVIDIA GeForce RTX 5060 Ti).

View 1
View 2
Output
Motorcycle, Middlebury 2014 stereo (left and right view, downscaled 4x), view 1
View 1 · 100%

Input: Motorcycle, Middlebury 2014 stereo (left and right view, downscaled 4x), Scharstein et al., Middlebury College (research dataset, no licence stated).