YOLO11-pose
Ultralytics' 2024 YOLO for multi-person pose with COCO's 17 keypoints, n to x, AGPL-3.0 or a paid Enterprise license.
Conditional for size n: code AGPL-3.0-only, weights AGPL-3.0-only. Measured on our RTX 4090: 14.2 ms median, 574 MiB VRAM peak. PyTorch weights, 6 MB.
Conditional for size s: code AGPL-3.0-only, weights AGPL-3.0-only. Measured on our RTX 4090: 14.3 ms median, 638 MiB VRAM peak. PyTorch weights, 19 MB.
Conditional for size m: code AGPL-3.0-only, weights AGPL-3.0-only. Measured on our RTX 4090: 15.1 ms median, 714 MiB VRAM peak. PyTorch weights, 40 MB.
Conditional for size l: code AGPL-3.0-only, weights AGPL-3.0-only. Measured on our RTX 4090: 18.8 ms median, 770 MiB VRAM peak. PyTorch weights, 51 MB.
Conditional for size x: code AGPL-3.0-only, weights AGPL-3.0-only. Measured on our RTX 4090: 18.7 ms median, 1.1 GiB VRAM peak. PyTorch weights, 113 MB.
License
Conditional
Conditional
Conditional
Conditional
Conditional
- Code
- AGPL-3.0-onlychecked 3 Nov 2026
- Weights
- AGPL-3.0-onlyall sizes, checked 3 Nov 2026
AGPL-3.0: shipping it in a product means releasing your source, unless you buy an Ultralytics Enterprise license.
AGPL-3.0: shipping it in a product means releasing your source, unless you buy an Ultralytics Enterprise license.
AGPL-3.0: shipping it in a product means releasing your source, unless you buy an Ultralytics Enterprise license.
AGPL-3.0: shipping it in a product means releasing your source, unless you buy an Ultralytics Enterprise license.
AGPL-3.0: shipping it in a product means releasing your source, unless you buy an Ultralytics Enterprise license.
On our RTX 4090, measured
14.2ms median
p90 14.8 ms, fp16, batch 1, 640x448
- VRAM peak
- 574 MiB
- Status
- Runs as is
Environment, 3 Nov 2026
NVIDIA GeForce RTX 4090, driver 580.173.02, CUDA 13.0, torch 2.14.1+cu130, Python 3.12.9, ultralytics==8.4.189. Baseline 801 MiB. 10 warm-up + 100 timed. Scope: predict(): pre-processing, forward pass and post-processing; input file decoded beforehand.
Accuracy
Not yet scored
Reported50 COCO mAPpose 50-95 on COCO val2017 keypointsSource: Ultralytics YOLO11 docs, pose performance table (mAPpose 50-95, docs/en/models/yolo11.md at commit 4fb6485b)
14.3ms median
p90 14.8 ms, fp16, batch 1, 640x448
- VRAM peak
- 638 MiB
- Status
- Runs as is
Environment, 3 Nov 2026
NVIDIA GeForce RTX 4090, driver 580.173.02, CUDA 13.0, torch 2.14.1+cu130, Python 3.12.9, ultralytics==8.4.189. Baseline 801 MiB. 10 warm-up + 100 timed. Scope: predict(): pre-processing, forward pass and post-processing; input file decoded beforehand.
Accuracy
Not yet scored
Reported58.9 COCO mAPpose 50-95 on COCO val2017 keypointsSource: Ultralytics YOLO11 docs, pose performance table (mAPpose 50-95, docs/en/models/yolo11.md at commit 4fb6485b)
15.1ms median
p90 16.1 ms, fp16, batch 1, 640x448
- VRAM peak
- 714 MiB
- Status
- Runs as is
Environment, 3 Nov 2026
NVIDIA GeForce RTX 4090, driver 580.173.02, CUDA 13.0, torch 2.14.1+cu130, Python 3.12.9, ultralytics==8.4.189. Baseline 801 MiB. 10 warm-up + 100 timed. Scope: predict(): pre-processing, forward pass and post-processing; input file decoded beforehand.
Accuracy
Not yet scored
Reported64.9 COCO mAPpose 50-95 on COCO val2017 keypointsSource: Ultralytics YOLO11 docs, pose performance table (mAPpose 50-95, docs/en/models/yolo11.md at commit 4fb6485b)
18.8ms median
p90 19.3 ms, fp16, batch 1, 640x448
- VRAM peak
- 770 MiB
- Status
- Runs as is
Environment, 3 Nov 2026
NVIDIA GeForce RTX 4090, driver 580.173.02, CUDA 13.0, torch 2.14.1+cu130, Python 3.12.9, ultralytics==8.4.189. Baseline 801 MiB. 10 warm-up + 100 timed. Scope: predict(): pre-processing, forward pass and post-processing; input file decoded beforehand.
Accuracy
Not yet scored
Reported66.1 COCO mAPpose 50-95 on COCO val2017 keypointsSource: Ultralytics YOLO11 docs, pose performance table (mAPpose 50-95, docs/en/models/yolo11.md at commit 4fb6485b)
18.7ms median
p90 19.2 ms, fp16, batch 1, 640x448
- VRAM peak
- 1.1 GiB
- Status
- Runs as is
Environment, 3 Nov 2026
NVIDIA GeForce RTX 4090, driver 580.173.02, CUDA 13.0, torch 2.14.1+cu130, Python 3.12.9, ultralytics==8.4.189. Baseline 801 MiB. 10 warm-up + 100 timed. Scope: predict(): pre-processing, forward pass and post-processing; input file decoded beforehand.
Accuracy
Not yet scored
Reported69.5 COCO mAPpose 50-95 on COCO val2017 keypointsSource: Ultralytics YOLO11 docs, pose performance table (mAPpose 50-95, docs/en/models/yolo11.md at commit 4fb6485b)
Weights
Outputs on the shared test photos
Same photos and one score threshold for every model of this task. Failures are marked.
Sizes
| Size | Latency | VRAM | Accuracy | Weights | Runs | License |
|---|---|---|---|---|---|---|
| n | 14.2 ms640x448 | 574 MiB | Not yet scored50 COCO mAPpose 50-95 paper | 6 MB | As is | Conditional |
| s | 14.3 ms640x448 | 638 MiB | Not yet scored58.9 COCO mAPpose 50-95 paper | 19 MB | As is | Conditional |
| m | 15.1 ms640x448 | 714 MiB | Not yet scored64.9 COCO mAPpose 50-95 paper | 40 MB | As is | Conditional |
| l | 18.8 ms640x448 | 770 MiB | Not yet scored66.1 COCO mAPpose 50-95 paper | 51 MB | As is | Conditional |
| x | 18.7 ms640x448 | 1.1 GiB | Not yet scored69.5 COCO mAPpose 50-95 paper | 113 MB | As is | Conditional |