watsonx.data pricing

The only hybrid, open data lakehouse for enterprise AI and analytics.

Start your free trial See the pricing
Power Gen AI

Scale and automate the preparation of structured and unstructured data to build more reliable generative AI applications.
Open and hybrid by design

Access data and deploy across on-premises, in the cloud, and in hybrid environments.
Optimize price performance for any workload

Select from multiple, fit-for-purpose query engines to optimize your workloads for the right balance of cost and performance.
Enterprise-grade security and governance

Secure, compliant, and audit-ready with enterprise-grade governance, lineage tracking, version control, access policies and more.
Choose from multiple payment options
Lite Getting started

Start with 500 free resource units to begin exploring today.

 Start your free trial SaaS Pay-As-You-Go - Recommended Enterprise production use

No upfront commitment. Pay only for what you use.

 Buy Now - Pay per use SaaS Subscription Enterprise production use

Maximize your budget with a set subscription package.

 Buy now - Subscription

*Prices shown are indicative, may vary by country, exclude any applicable taxes and duties, and are subject to product offering availability in a locale.

Only pay for what you use

Our managed service offering provides consumption-based pricing that scales with your business, ensuring that you only pay for what you use.*
Small On-Demand

Resources are dynamically provisioned and automatically scaled up to the maximum cores available. Pricing is flexible based on 0.2 RU/hr/vCPU core +
0.02 RU/hr/GB

 Medium (Balanced) at 2.0 RUs/hr

16 vCPUs, 64 GiB RAM, plus ~ 900 GB of secondary storage

 Medium (Storage Optimized) at 2.8 RUs/hr

16 vCPUs, 128 GiB RAM, ~1600 GB of secondary storage.

 Large (Balanced) at 3.69 RUs/hr

32 vCPUs, 256 GiB RAM, plus ~ 1600 GB of secondary storage.

 Large (Storage Optimized) at 5.5 RUs/hr

32 vCPUs, 256 GiB RAM, ~3200 GB of secondary storage.
Starter at 1.25 RUs/hr

1 million vectors, Index Type- IVF_FLAT,  Index Parameters – 64, Segment size (1024), Dimensions (384 to match IBM models) [Default size for trial].

 Small at 2.75 RUs/hr

10 million vectors, Index Type- IVF_FLAT,  Index Parameters – 64, Segment size (1024), Dimensions (at 384).

 Medium at 9.75 RUs/hr

50 million vectors, Index Type- IVF_FLAT,  Index Parameters – 64, Segment size (1024), Dimensions (at 384).

 Large at 16.50 RUs/hr

100 million vectors, Index Type- IVF_FLAT,  Index Parameters – 64, Segment size (1024), Dimensions (at 384).
Starter at 2.00 RUs/hr

1 Head node, 1 worker each with 1 × 2400GB NVMe, virtio (lakehouse usable), Up to 24 Gbps.

 Small at 11.20 RUs/hr

1 Head nodes, 3 worker nodes 1x 2400 virtio (reserved for container images), 1x 2400 virtio (lakehouse usable), Up to 24 Gbps.

 Medium at 19.60 RUs/hr

1 Head nodes, 6 worker nodes 1x 2400 virtio (reserved for container images), 1x 2400 virtio (lakehouse usable), Up to 24 Gbps.

 Large at 36.40 RUs/hr

1 Head nodes, 12 worker nodes 1x 2400 virtio (reserved for container images), 1x 2400 virtio (lakehouse usable), Up to 24 Gbps.
Starter at 0.7 RUs/hr

Optimized for ~16 GB storage; supports up to 80 reads/s, 80 writes/s, and 100 ops/s peak throughput.

 Extra small at 2.3 RUs/hr

1 Head nodes, 3 worker nodes 1x 2400 virtio (reserved for container images), 1x 2400 virtio (lakehouse usable), Up to 24 Gbps.

  Small at 8.6 RUs/hr

Optimized for ~1 TB storage; supports up to 1,000 reads/s, 1,000 writes/s, and 1,250 ops/s peak throughput.

 Medium at 17.7 RUs/hr

Optimized for ~2 TB storage; supports up to 5,000 reads/s, 5,000 writes/s, and 40,000 ops/s peak throughput.

 Large at 68.2 RUs/hr

Optimized for ~10 TB storage; supports up to 20,000 reads/s, 20,000 writes/s, and 160,000 ops/s peak throughput.

 Extra Large at 167.5 RUs/hr

Optimized for ~20 TB storage; supports up to 50,000 reads/s, 50,000 writes/s, and 500,000 ops/s peak throughput.
Small On-Demand

Resources are dynamically provisioned and automatically scaled up to the maximum cores available. Pricing is flexible based on 0.2 RU/hr/vCPU core +
0.02 RU/hr/GB

 Medium (Balanced) at 2.0 RUs/hr

16 vCPUs, 64 GiB RAM, plus ~ 900 GB of secondary storage

 Medium (Storage Optimized) at 2.8 RUs/hr

16 vCPUs, 128 GiB RAM, ~1600 GB of secondary storage.

 Large (Balanced) at 3.69 RUs/hr

32 vCPUs, 256 GiB RAM, plus ~ 1600 GB of secondary storage.

 Large (Storage Optimized) at 5.5 RUs/hr

32 vCPUs, 256 GiB RAM, ~3200 GB of secondary storage.
Starter at 1.25 RUs/hr

1 million vectors, Index Type- IVF_FLAT,  Index Parameters – 64, Segment size (1024), Dimensions (384 to match IBM models) [Default size for trial].

 Small at 2.75 RUs/hr

10 million vectors, Index Type- IVF_FLAT,  Index Parameters – 64, Segment size (1024), Dimensions (at 384).

 Medium at 9.75 RUs/hr

50 million vectors, Index Type- IVF_FLAT,  Index Parameters – 64, Segment size (1024), Dimensions (at 384).

 Large at 16.50 RUs/hr

100 million vectors, Index Type- IVF_FLAT,  Index Parameters – 64, Segment size (1024), Dimensions (at 384).
Starter at 2.00 RUs/hr

1 Head node, 1 worker each with 1 × 2400GB NVMe, virtio (lakehouse usable), Up to 24 Gbps.

 Small at 11.20 RUs/hr

1 Head nodes, 3 worker nodes 1x 2400 virtio (reserved for container images), 1x 2400 virtio (lakehouse usable), Up to 24 Gbps.

 Medium at 19.60 RUs/hr

1 Head nodes, 6 worker nodes 1x 2400 virtio (reserved for container images), 1x 2400 virtio (lakehouse usable), Up to 24 Gbps.

 Large at 36.40 RUs/hr

1 Head nodes, 12 worker nodes 1x 2400 virtio (reserved for container images), 1x 2400 virtio (lakehouse usable), Up to 24 Gbps.
Starter at 0.7 RUs/hr

Optimized for ~16 GB storage; supports up to 80 reads/s, 80 writes/s, and 100 ops/s peak throughput.

 Extra small at 2.3 RUs/hr

1 Head nodes, 3 worker nodes 1x 2400 virtio (reserved for container images), 1x 2400 virtio (lakehouse usable), Up to 24 Gbps.

  Small at 8.6 RUs/hr

Optimized for ~1 TB storage; supports up to 1,000 reads/s, 1,000 writes/s, and 1,250 ops/s peak throughput.

 Medium at 17.7 RUs/hr

Optimized for ~2 TB storage; supports up to 5,000 reads/s, 5,000 writes/s, and 40,000 ops/s peak throughput.

 Large at 68.2 RUs/hr

Optimized for ~10 TB storage; supports up to 20,000 reads/s, 20,000 writes/s, and 160,000 ops/s peak throughput.

 Extra Large at 167.5 RUs/hr

Optimized for ~20 TB storage; supports up to 50,000 reads/s, 50,000 writes/s, and 500,000 ops/s peak throughput.

*Prices shown are indicative, may vary by country, exclude any applicable taxes and duties, and are subject to product offering availability in a locale. Pricing includes compute costs only and does not include the core support services cost per account of 3.00 RUs/hr. Usage/metering is tracked at a per-minute level. Learn more here
Resource units (RU) is our unit of measure for compute consumption and is available at a list price of USD 1 per RU, metered on a per-second basis with a 1-minute minimum charge

Find the right fit with flexible deployment options

SaaS Managed Service Offering

Consumption-based pricing that lets you pay only for what you use and eliminate idle spend by pausing resources. This option is easy-to-setup and ready to deploy in minutes, ensuring fast time-to-value for your business.
Bring Your Own Cloud (BYOC)

Optimize your cloud spend by paying only for watsonx.data software while leveraging your infrastructure and and existing cloud provider credits. This option ensures infrastructure sovereignty by giving you control over data residency and security within your own VPC.

Contact sales
On-Prem Software Offering

Entitlement-based pricing allows you to effectively and predictably manage your costs. Leverage your existing investments in IBM Power and Fusion HCI to optimize performance.

Contact sales

FAQ

 

Yes. You can run IBM watsonx.data as a fully managed multi-cloud service on IBM Cloud, AWS or in your own on-premises environment. This flexibility lets you choose what fits your architecture and compliance needs. 

 

You can explore simple tiered plans designed to help you get started fast: a Free Trial to experiment, Enterprise plan for scaling your data lakehouse.

You can buy directly from the IBM Cloud Catalog or through AWS Marketplace

Take the next step

It’s time to make the most of your data for gen AI.

 Start your free trial Contact sales
Footnotes

Unless otherwise specified under Software pricing, all features, capabilities and potential updates refer exclusively to SaaS. IBM makes no representation that SaaS and software features and capabilities will be the same.

*Prices shown are indicative, may vary by country, exclude any applicable taxes and duties, and are subject to product offering availability in a locale. Pricing includes compute costs only and does not include the core support services cost per account of 3.00 RUs/hr. Usage/metering is tracked at a per-minute level. Learn more here
Resource units (RU) is our unit of measure for compute consumption and is available at a list price of USD 1 per RU, metered on a per-second basis with a 1-minute minimum charge

For full details on payment methods and purchase eligibility by country, visit this page.