ABOUT TENSORQUAY / SINGAPORE

Local AI.
Built around your work.

TensorQuay PTE. LTD. is a Singapore company providing customer-owned AI workstations, local model deployment and workflow configuration for individuals and small teams.

One workstation. The complete AI setup.

TQ–32 · Currently not available

S$15,615 for a workstation with RTX 5090, local model and agent setup, testing, documentation and a guided handover.

TQ–32 and ordering options

TQ–A128 · Available to order

S$8,490 for an agent server for teams (16 cores / 32 threads, 128 GB RAM, no GPU needed) with TensorQuay Agent, our open, on-premise agent platform (pre-release), set up for your team. Delivered within 1 week, within Singapore.

TQ–A128 specifications

TQ–192 · Proposed

A proposed system with two RTX PRO 6000 Max-Q GPUs and 192 GB of GPU memory, for larger local models and team workloads. Final pricing, supply and the supported setup are confirmed per quotation. It is available by enquiry only.

Proposed TQ–192 specifications

Every TQ–32 includes the agreed AI setup. There is no separate hardware-only or Ubuntu-only edition. The supplied interface, supported tools and acceptance tasks are agreed for your deployment. Exact component supply and complete-system qualification remain subject to confirmation; see the hardware specification and its qualifications.

Deployment, adaptation and support.

We also scope local AI deployments on compatible existing infrastructure. Work can be carried out remotely or on site within Singapore, with access arrangements and a quotation agreed first. On-site installation is not automatically included in ordinary hardware delivery.

Fine-tuning and post-training require a separately agreed task, dataset, model licence, compute plan and evaluation. Ongoing maintenance and new workflows also have their own scope. Explore our local AI services and the deployment process.

Evidence you can inspect.

Our Qwen / RTX 5090 test notes include recorded excerpts, completion outcomes and failures from internal tests on rented cloud hardware. They establish observations under those test conditions, not the performance of a finished workstation or a guaranteed customer count.

The local AI guides explain how to evaluate a workflow, choose tools and understand data flow. We describe the limits alongside the results. No NVIDIA partnership or endorsement is claimed.

Company identity and contact.

Registered company: TensorQuay PTE. LTD.
Singapore UEN: 202643088C
Contact: hello@tensorquay.com
Official profile: TensorQuay on X

To discuss a project, describe your task, users and current setup. The website AI assistant uses a third-party cloud service; it is separate from local AI on a purchased workstation. You can choose human support in the website chat.

Email our team