Sold by: HolboxAI
Deployed on AWS
Private OpenAI-compatible vLLM endpoint serving Qwen3-4B on a GPU EC2 instance in your account. Function/tool calling and 32K context. Returns an endpoint URL and API key.
Overview
Qwen 3.5 4B Inference Endpoint deploys a private, OpenAI-compatible large language model endpoint into your own AWS account. It serves the Qwen3.5 4B model with vLLM on a single GPU EC2 instance, so your prompts and data never leave your environment.
The endpoint is OpenAI API compatible (v1/chat/completions, v1/models) and supports function/tool calling and a 32K token context window, making it a drop-in local model for IDE coding agents such as Cursor, Cline, and Continue, as well as your own apps.
The AMI is self-contained: on first boot it starts vLLM, generates an API key, and serves on port 8000. First boot downloads the model (about 18 to 25 minutes). You pay only for the EC2 GPU instance while it runs; terminate the instance to stop all charges.
Highlights
- Private OpenAI-compatible Qwen3.5 4B endpoint that runs entirely in your own AWS account - prompts and data never leave your environment.
- Function/tool calling and 32K token context out of the box - works as a local model for IDE agents like Cursor, Cline, and Continue.
- Self-contained AMI: boots vLLM automatically, generates an API key, and serves on port 8000. Pay only for the EC2 GPU while it runs.
Details
Sold by
Categories
Delivery method
Delivery option
64-bit (x86) Amazon Machine Image (AMI)
Latest version
Operating system
Ubuntu 22.04
Deployed on AWS
New
Introducing multi-product solutions
You can now purchase comprehensive solutions tailored to use cases and industries.
Features and programs
Financing for AWS Marketplace purchases
AWS Marketplace now accepts line of credit payments through the PNC Vendor Finance program. This program is available to select AWS customers in the US, excluding NV, NC, ND, TN, & VT.
Pricing
This product is available free of charge. Free subscriptions have no end date and may be canceled any time.
Additional AWS infrastructure costs may apply. Use the AWS Pricing Calculator to estimate your infrastructure costs.
Vendor refund policy
This product is offered free of charge by Holbox AI. There are no software charges, and therefore nothing to refund through this listing.
You remain responsible for the AWS infrastructure costs of running it - the EC2 instance and its EBS root volume. Those are billed by AWS, not by Holbox AI, and are not refundable through this listing. Stop or terminate instances when they are idle. For questions about AWS charges, contact AWS Support or see https://aws.amazon.com/marketplace/help/
Legal
Vendor terms and conditions
Upon subscribing to this product, you must acknowledge and agree to the terms and conditions outlined in the vendor's End User License Agreement (EULA) .
Content disclaimer
Vendors are responsible for their product descriptions and other product content. AWS does not warrant that vendors' product descriptions or other product content are accurate, complete, reliable, current, or error-free.
Delivery details
64-bit (x86) Amazon Machine Image (AMI)
Amazon Machine Image (AMI)
An AMI is a virtual image that provides the information required to launch an instance. Amazon EC2 (Elastic Compute Cloud) instances are virtual servers on which you can run your applications and workloads, offering varying combinations of CPU, memory, storage, and networking resources. You can launch as many instances from as many different AMIs as you need.
Version release notes
Initial release. Qwen-3.5 4B on vLLM. Hermes tool/function calling, 32K context, OpenAI-compatible API on port 8000
Additional details
Usage instructions
-
Launch the product on a recommended GPU instance (g5.xlarge or larger) in a public subnet. In the security group, allow inbound TCP 8000 from your IP or CIDR.
-
First boot downloads the model and starts vLLM (about 18 to 25 minutes). The service starts automatically - no SSH or extra configuration is required.
-
Retrieve the endpoint URL and API key without SSH:
- Serial console: EC2 console > select the instance > Actions > Monitor and troubleshoot > Get system log. Look for the "Qwen4B endpoint" banner with the Endpoint and API key.
- Optional (SSM Parameter Store): if you attach an IAM instance profile that grants ssm:PutParameter on /qwen4b/*, the boot script also publishes the endpoint and key to /qwen4b/INSTANCE_ID/endpoint and /qwen4b/INSTANCE_ID/api_key.
-
Call the endpoint (model name is qwen4b): curl http://PUBLIC_IP:8000/v1/chat/completions
-H "Authorization: Bearer YOUR_API_KEY"
-H "Content-Type: application/json"
-d '{"model":"qwen4b","messages":[{"role":"user","content":"Hello"}]}' -
Tool/function calling: pass an OpenAI-style "tools" array; the endpoint returns structured tool_calls (Hermes parser). Context window is 32768 tokens.
-
To stop all charges, terminate the instance (or delete the CloudFormation stack if you launched via a template).
Support
Vendor support
Support email: support@holbox.ai
Support description: Support is provided via email for product installation, deployment, configuration, and troubleshooting. Assistance includes AMI startup problems, endpoint availability, and product-specific questions. Please include your AWS Region, CloudFormation stack events, and relevant logs when contacting support to help us investigate issues more efficiently.
AWS infrastructure support
AWS Support is a one-on-one, fast-response support channel that is staffed 24x7x365 with experienced and technical support engineers. The service helps customers of all sizes and technical abilities to successfully utilize the products and features provided by Amazon Web Services.
Similar products
Best-in-class legal enrichment, reranking, and embedding models, optimized for legal knowledge graphs and legal RAG.
A state-of-the-art reranking model optimized for legal RAG, research, and classification, ranked first on Legal RAG Bench and the Massive Legal Embedding Benchmark (MLEB).
Nemotron-3-Nano-30B-A3B is a large language model (LLM) trained from scratch by NVIDIA, and designed as a unified model for both reasoning and non-reasoning tasks. It responds to user queries and tasks by first generating a reasoning trace and then concluding with a final response.
Built on openSUSE Linux, this product provides private AI using the Qwen 3 model with 14 billion parameters. MultiCortex HPC (High-Performance Computing) allows you to boost your AI's response quality. This is a plug-and-play, low-cost product with no token fees.
Customer reviews
No customer reviews yet
Be the first to review this product . We've partnered with PeerSpot to gather customer feedback. You can share your experience by writing or recording a review, or scheduling a call with a PeerSpot analyst.