owlv2-base-patch16-ensemble by Google: Open-Weight Model

Runs On

What it takes to serve owlv2-base-patch16-ensemble (155M 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	0.4 GB	1x MI300X (192 GB) Vultr	$1.85	1x H100 $1.99 · 1x MI325X $2.00
8-bit	0.2 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 owlv2-base-patch16-ensemble

Point a camera at a construction site, ask for hard hats and forklifts by name, and get boxes back without training a detector for either. Google's owlv2-base-patch16-ensemble does that: text-conditioned detection on a CLIP backbone at 155M parameters. Memory is not the constraint: 0.4 GB at 16-bit, 0.2 GB at 8-bit, 0.1 GB at 4-bit, with the cheapest listed setup one MI300X at 192 GB for $1.85 per hour on-demand. Stored weights are float32, so converting is on you.

Apache 2.0 applies, with commercial use, redistribution and a patent grant. The weights were released October 13, 2023 and last touched October 31, 2024, the paper behind them is arXiv:2306.09683, and there is no context length figure, since inputs are images plus short text queries. Files ship in safetensors and PyTorch. With no Index host prices, per-image cost is frames per hour through one rented card.

SAVRN Research, 2026-09-18

Model Card

By Google, published under apache-2.0, revision cfd3195ba4ea.

Model Card: OWLv2

Model Details

The OWLv2 model (short for Open-World Localization) was proposed in Scaling Open-Vocabulary Object Detection by Matthias Minderer, Alexey Gritsenko, Neil Houlsby. OWLv2, like OWL-ViT, is a zero-shot text-conditioned object detection model that can be used to query an image with one or multiple text queries.

Read the full model card (595 words)

Configuration

Architecture: Owlv2ForObjectDetection
Stored precision: float32
Model type: owlv2

Identity and Version

Repository: google/owlv2-base-patch16-ensemble
Publisher: Google
Task: Zero shot object detection
Modality: Other
Library: transformers
Parameters: 155M parameters
Languages: Not stated by the source
Revision: cfd3195ba4ea9592eec887ded089f4c08eff231d
First published: 2023-10-13
Last updated: 2024-10-31

Files and Weights

11 files, 1.2 GB in total. The weights are 2 files totalling 1.2 GB in bin, safetensors.

Weights2 files · 1.2 GB

Configuration4 files · 1.0 KB

Tokenizer3 files · 1.6 MB

Documentation1 file · 4.8 KB

Repository1 file · 1.5 KB

Every file

File	Type	Size	SHA-256
model.safetensors	Weights	619.9 MB	e1e130b9e404
pytorch_model.bin	Weights	620.0 MB	69feda8b53b1
added_tokens.json	Configuration	67 B	—
config.json	Configuration	414 B	—
preprocessor_config.json	Configuration	425 B	—
special_tokens_map.json	Configuration	121 B	—
README.md	Documentation	4.8 KB	—
.gitattributes	Repository	1.5 KB	—
merges.txt	Tokenizer	524.6 KB	—
tokenizer_config.json	Tokenizer	1.1 KB	—
vocab.json	Tokenizer	1.1 MB	—

License and Download

License: apache-2.0
Access: Open weights, no gate
Download size: 1.2 GB

Download from Google

Released by Google through its official repository on Hugging Face. Read the license.

Built From

Described by arXiv:2306.09683

Memory Requirements

Precision	Weights in memory
As published	1.2 GB
16-bit	0.3 GB
8-bit	0.2 GB
4-bit	0.1 GB

Weights only, from the published parameter count; the key-value cache and runtime add to this.

Questions About owlv2-base-patch16-ensemble

How much GPU memory does owlv2-base-patch16-ensemble need?

About 0.4 GB at 16-bit and 0.1 GB at 4-bit: the weights (155M parameters) plus a working margin. A long context needs more.

What is the cheapest GPU to run owlv2-base-patch16-ensemble 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 owlv2-base-patch16-ensemble commercially?

Yes. owlv2-base-patch16-ensemble 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.

Similar Models

Model · Zero shot object detection

grounding-dino-tiny

IDEA-Research

The Grounding DINO model was proposed in Grounding DINO: Marrying DINO with Grounded Pre-Training for Open-Set Object Detection by Shilong Liu, Zhaoyang Zeng, Tianhe Ren, Feng Li, Hao Zhang, Jie Yang, Chunyuan Li, Jianwei Yang, Hang Su, Jun Zhu, Lei Zhang. Grounding DINO extends a closed-set object detection model with a text encoder, enabling open-set object detection. The model achieves remarkable results, such as 52.5 AP on COCO zero-shot. alt="drawing" width="600"/> You can use the raw model for zero-shot object detection (the task of detecting things in an image out-of-the-box without labeled data). Here's how to use the model for zero-shot object detection

Open weights apache-2.0 172M parameters transformers

View model

Model · Zero shot object detection

grounding-dino-base

IDEA-Research

The Grounding DINO model was proposed in Grounding DINO: Marrying DINO with Grounded Pre-Training for Open-Set Object Detection by Shilong Liu, Zhaoyang Zeng, Tianhe Ren, Feng Li, Hao Zhang, Jie Yang, Chunyuan Li, Jianwei Yang, Hang Su, Jun Zhu, Lei Zhang. Grounding DINO extends a closed-set object detection model with a text encoder, enabling open-set object detection. The model achieves remarkable results, such as 52.5 AP on COCO zero-shot. alt="drawing" width="600"/> You can use the raw model for zero-shot object detection (the task of detecting things in an image out-of-the-box without labeled data). Here's how to use the model for zero-shot object detection

Open weights apache-2.0 233M parameters transformers

View model