Pricing

Transparent pricing, flexible plans, pay as you go — you only pay for what you actually use.

Model API

Access 7+ leading models with token-based billing and elastic scaling.

InputBilled per input token

OutputBilled per output token

PrecisionFP8 / FP16 selectable

Minimum spendNo minimum spend

BillingPay-as-you-go, no minimum spend

Rent the full NVIDIA and Huawei Ascend GPU lineup on demand, provisioned in seconds.

RTX 4090Top pick for inference · great value

A100Training workhorse · 80G VRAM

H100Flagship training · blazing speed

Ascend 910BDomestic compute · self-controlled

MonthlyBetter value for long-term use

Structured courses and official certification, from beginner to expert.

Online coursesLearn anytime

Offline lecturesCitywide tours

Enterprise trainingTailor-made

Official certificationSkill endorsement

Deploy full AI capabilities in your own data center — your data never leaves the premises.

StandardSingle node to hundreds of cards

AdvancedMulti-region clusters

ComplianceMLPS Level 3 / national crypto

Operations7×24 in-house support

How do I get started with the API?

Sign up, top up, then create an API Key in the console. Usage is deducted from your balance — no minimum spend, no monthly fee.

How is GPU compute billed?

Billed per second of actual usage. We pre-freeze 24 hours of fees when an instance is created, and refund the remainder when you stop it.

Do you support private deployment?

Yes. We offer enterprise private deployment so your data stays on-premises, meeting compliance needs in finance, healthcare and more.

Can I try courses for free?

Yes. The platform offers many free introductory courses, and paid courses include previewable chapters.

How do I contact sales?

Click “Book a consultation” on the page or email sales@xable.com — an advisor will reach out within 2 hours.

Whether you are an individual or an enterprise, there is an option for your AI needs.