Private LLM and RAG server rental — ASUS GX10
Rent ASUS Ascent GX10 to evaluate your own AI assistant, RAG search or open-model API. Choose the model and inference settings for the application. One node provides 128 GB of unified CPU and GPU memory; a pair can separate experiments or workloads across two systems.
An LLM alongside your application
Running your own runtime lets you choose the model, system instructions, context and application logging policy. Before uploading working documents, agree on node access, network routing, data storage and deletion. Confirm access arrangements and server location separately: the label “private LLM” does not establish these terms.
RAG over your documents
Prepare a small document set and questions with known answers. Evaluate retrieval quality separately from the model’s answer quality. Weights, context, runtime and working data must fit in memory; the index and embedding service may need additional resources. Test the full path from query to response.
Evaluating your own LLM API
Record the engine version, model format and maximum context. For an API, measure first-token latency, response duration, queuing and behavior at peak concurrency. Check ARM64 and GB10 compatibility before a longer run. Save results with their settings so the evaluation can be reproduced.
A configuration for your workload
Request one ASUS GX10 for a pilot application or two nodes for separate models and tests. Compatible software distributes a shared workload across the two 128 GB systems. Include the model, evaluation data, access requirements and duration in your request. Rates are listed on the configuration pages.
ASUS GX10 rental rate
One hourly rate applies to the selected rental duration. The calculator multiplies it by the full rental period.
The rate is for one ASUS Ascent GX10 with 128 GB of memory shared by the CPU and GPU.
Choose ASUS for your model
Share the model, quantization, context and dates. Choose one node or a pair and agree on a configuration for your workload.
Models for ASUS GX10ASUS GX10 rental
Before you rent
Can I run a model without an external LLM API?
Consider open weights and your own runtime on GX10. Check the model licence, dependencies and whether the application calls external services.
Does rental guarantee document isolation?
Access, network, logging and deletion requirements must be agreed before uploading documents. This page describes a workload evaluation, rather than confirmed data protection terms.
One GX10 or two for RAG?
For a pilot, start by evaluating the complete pipeline on one node. Two systems can separate workloads or models; the choice depends on memory, load and application architecture.
What should I include in the request?
Include the model, context size, example documents and questions, expected concurrency, access requirements and dates. An initial description can omit confidential data.