Object storage
Object storage built for AI and ML
S3 compatible object storage designed for fast, reliable access to training data, model outputs and large-scale AI workflows.
Storage that keeps up with your models
Store, access and scale your AI data with ease. From training data to model outputs, object storage supports every stage of the AI lifecycle.
Scalable by design
Handle growing datasets with ease. Scale storage on demand as your models and data expand
Close to your compute
Run storage alongside your compute to reduce latency and speed up training and inference.
Scale without friction
Grow your datasets without reconfiguring or hitting performance limits. Scale as you train.
Made for modern AI teams
Adaptable to every step of the AI lifecycle, from prototyping to production.
Transparent pricing for performance at scale
Store your data close to your compute, with pricing that reflects performance and scale. No egress fees. No hidden costs. Just reliable storage built for AI.
| Data center | Cost per GB | |
|---|---|---|
ng-lagos-1 | $0.01/mo | |
us-losangeles-1 | $0.01/mo | |
us-newyork-1 | $0.01/mo | |
us-santaclara-1 | $0.01/mo | |
no-kristiansand-1 | $0.05/mo | |
gb-bournemouth-1 | $0.08/mo | |
ca-montreal-3 | $0.09/mo | |
se-stockholm-1 | $0.09/mo | |
au-melbourne-1 | - | |
ca-montreal-2 | - | |
fi-tampere-1 | - | |
gb-manchester-1 | - | |
in-hyderabad-1 | - | |
us-carlsbad-1 | - | |
za-centurion-1 | - |
Integrate seamlessly with virtual machines
Deploy cost-effective virtual machines on our platform for a wide range of workloads.
Control both data and infra with bare metal
Give yourself more control over your infrastructure with our on-demand bare metal
Built for real-world AI workloads
From training foundation models to powering real-time inference, our object storage fits seamlessly into your AI stack.
Training LLMs
Store and stream massive datasets used in large language model training.
Fine-tuning models
Keep your fine-tuning data and checkpoints close to compute for faster iterations.
Inference at scale
Feed inference jobs at low latency with reliable data access across regions.
Model artefact storage
Manage and version model outputs, logs and evaluation results in one place.
An essential part of your AI infrastructure
Modern AI workloads depend on fast, reliable access to data. Our object storage is built to handle unstructured datasets at scale, making it a critical layer for training, inference and collaboration.
Model training
Store large datasets close to your compute for efficient model training with deep learning frameworks.
Inference Pipelines
Serve input and output data at low latency to keep inference fast, whether in real time or batch.
Experiment tracking
Persist logs, checkpoints and artefacts to compare models and ensure repeatable results.
Collaboration
Share datasets and results across teams and regions with built-in replication and version control.
Frequently asked questions
Find the resources to empower your AI journey