Open-weight model · Keypoint detection
vitpose-plus-base
by University of Sydney usyd-community/vitpose-plus-base
ViTPose: Simple Vision Transformer Baselines for Human Pose Estimation and ViTPose+: Vision Transformer Foundation Model for Generic Body Pose Estimation. It obtains 81.1 AP on MS COCO Keypoint test-dev set.
Runs On
What it takes to serve vitpose-plus-base (125M parameters): the memory its weights need at each precision, and the cheapest way to rent enough data-center GPUs to hold them.
| Precision | Weights | Memory needed | Cheapest setup | Per hour | Also fits |
|---|---|---|---|---|---|
| 16-bit | 0.3 GB | 0.3 GB | 1x MI300X (192 GB) Vultr |
$1.85 | 1x H100 $1.99 · 1x MI325X $2.00 |
| 8-bit | 0.1 GB | 0.2 GB | 1x MI300X (192 GB) Vultr |
$1.85 | 1x H100 $1.99 · 1x MI325X $2.00 |
| 4-bit | 0.1 GB | 0.1 GB | 1x MI300X (192 GB) Vultr |
$1.85 | 1x H100 $1.99 · 1x MI325X $2.00 |
Memory is the weights at that precision plus 20% for the runtime and a short context; a long context needs more. Prices are the lowest on-demand hourly rates in the SAVRN Index, read Sep 18, 2026.
SAVRN's Notes on vitpose-plus-base
Three tenths of a gigabyte is the whole memory budget for this keypoint detector at 16-bit precision, and it falls to 0.2 GB at 8-bit and 0.1 GB at 4-bit. The 125M parameters ship as float32 in five files, about 502 MB. Its job is human pose estimation: finding body keypoints in images. The cheapest slot on our board is one 192 GB MI300X at $1.85 per hour on-demand, which this model leaves almost idle, so run it as a sidecar on a card already doing other work.
Apache 2.0 permits commercial use, modification and redistribution, provided the license and copyright notices travel with the weights and significant changes are stated. There is no context length to check and no per-token host price on the SAVRN Index; pose estimation is billed by the GPU hour. Read arXiv 2204.12484 before committing; the weights date from January 8, 2025.
Model Card
By University of Sydney, published under apache-2.0, revision 92be54d7a29e.
Model Card for VitPose
ViTPose: Simple Vision Transformer Baselines for Human Pose Estimation and ViTPose+: Vision Transformer Foundation Model for Generic Body Pose Estimation. It obtains 81.1 AP on MS COCO Keypoint test-dev set.
Model Details
Configuration
- Architecture
- VitPoseForPoseEstimation
- Stored precision
- float32
- Model type
- vitpose
Identity and Version
- Repository
- usyd-community/vitpose-plus-base
- Publisher
- University of Sydney
- Task
- Keypoint detection
- Modality
- Other
- Library
- transformers
- Parameters
- 125M parameters
- Languages
- en
- Revision
- 92be54d7a29e42fad47b6e2ca01dd9e685a61e0d
- First published
- 2025-01-08
- Last updated
- 2025-01-08
Files and Weights
5 files, 501.6 MB in total. The weights are 1 file totalling 501.6 MB in safetensors.
Every file
| File | Type | Size | SHA-256 |
|---|---|---|---|
| model.safetensors | Weights | 501.6 MB | 640225e4a9dd |
| config.json | Configuration | 1.8 KB | — |
| preprocessor_config.json | Configuration | 363 B | — |
| README.md | Documentation | 11.5 KB | — |
| .gitattributes | Repository | 1.5 KB | — |
License and Download
- License
- apache-2.0
- Access
- Open weights, no gate
- Download size
- 501.6 MB
Released by University of Sydney through its official repository on Hugging Face. Read the license.
Built From
- Described by arXiv:2204.12484
Memory Requirements
| Precision | Weights in memory |
|---|---|
| As published | 501.6 MB |
| 16-bit | 0.3 GB |
| 8-bit | 0.1 GB |
| 4-bit | 0.1 GB |
Weights only, from the published parameter count; the key-value cache and runtime add to this.
Questions About vitpose-plus-base
How much GPU memory does vitpose-plus-base need?
About 0.3 GB at 16-bit and 0.1 GB at 4-bit: the weights (125M parameters) plus a working margin. A long context needs more.
What is the cheapest GPU to run vitpose-plus-base on?
At 16-bit, 1x MI300X from $1.85 an hour; at 4-bit, 1x MI300X from $1.85 an hour, at the lowest on-demand prices the SAVRN Index lists.
Can I use vitpose-plus-base commercially?
Yes. vitpose-plus-base is released under Apache License 2.0. The Apache License 2.0 is a permissive open-source license. It permits commercial use, modification and redistribution. It requires keeping the license and copyright notices and any NOTICE file, stating significant changes, and it includes an express patent grant from contributors.