A self-hosted, production-ready Qwen 3.8 27B Uncensored model deployed into your AWS environment with a single click. Because everything runs entirely within your private cloud, your data stays secure, isolated, and fully under your control. Best of all, unlimited tokens.
This is a self-hosted deployment of the Qwen 3.8 27B Uncensored large language model. It runs as a single GPU-powered EC2 instance allowing you to keep your data private and leverage unlimited tokens. Access is over HTTPS on port 443 with a per-instance API key; Ollama itself listens only on localhost. GPU instance types need EC2 quota for "Running On-Demand G and VT instances" (L-DB2E81BA), which new AWS accounts often have set to 0. The model is preloaded into GPU memory on boot, so the first prompt does not wait for a cold load. Once the instance is powered on, the model is ready in about 6 minutes. Highlights of the Qwen 3.8 Uncensored model include:
Abliterated (refusal-removed) dense 27B hybrid-attention vision-language model, with the vision tower and MTP speculative-decoding head preserved from Qwen 3.8 27B.
Native context window of 262,144 tokens.
Thinking mode is on by default and can be toggled per request; tool calling is supported.
Native vision encoder for image understanding, including on-image text (OCR).
Hybrid Gated DeltaNet and full-attention architecture (64 layers).
Capability on MMLU, MMLU-Pro, GSM8K, and CMMLU remains within 1.3 points of the official Qwen 3.8 27B FP8 base on the published card.
Released under the Apache 2.0 license, inherited from the base model. Intended for research, interpretability, red-teaming, and controlled evaluation; add your own safety and moderation layers before any end-user deployment.
Production-ready Qwen 3.8 27B Uncensored with HTTPS and API key
Operating system
Ubuntu 26.04
Deployed on AWS
Yes
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.
Try this product free for 5 days according to the free trial terms set by the vendor. Usage-based pricing is in effect for usage beyond the free trial terms. Your free trial gets automatically converted to a paid subscription when the trial ends, but may be canceled any time before that.
You pay by the hour for the EC2 instance that runs this model in your AWS account. Pricing is not based on tokens or requests. You pick from six GPU instance sizes across three families: g5, g6, and g6e. Within each family, the xlarge and 2xlarge options set the compute and memory you get. Larger instances carry higher hourly rates. Choose the instance that matches your workload, since the hourly fee scales with the instance you select.
Top-of-mind questions for buyers
What does one billing unit cover for each instance option?
Each unit is one hour that the chosen EC2 GPU instance runs in your AWS account. The model runs on that machine, so you pay for the compute, not for tokens or requests. One hour of an xlarge or 2xlarge instance counts as one unit.
Am I charged when the instance is stopped?
The hourly software fee applies only while the instance runs. If you stop the instance, the hourly charge stops too. Underlying AWS resources, such as attached storage, may still incur separate AWS fees while the instance is stopped.
How do the six instance options differ in what drives my cost?
Your cost depends on which instance you pick and how many hours it runs. You pay for one instance at a time, not a combination. The three families, g5, g6, and g6e, set the hardware. Within each, xlarge and 2xlarge set compute and memory, with higher rates for more resources.
Tell us how we can improve this page, or report an issue with this product.
Give us feedbackReport a problem with this product or seller
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.
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
HTTPS on port 443 with a per-instance API key (Bearer). Ollama now listens on 127.0.0.1:11434 only and OLLAMA_ORIGINS is no longer "*" (set SALIENT_ALLOWED_ORIGINS to allow browser apps). The baked model is read off EBS in parallel at boot, so it is ready in about 6 minutes instead of up to 90. The EC2 system log and the SSH login banner show the endpoint, API key, TLS fingerprint and example requests. Added instance types: g6.xlarge, g6.2xlarge, g5.xlarge, g5.2xlarge, g6e.xlarge.
Upgrading from an earlier version: clients that called http://:11434 must switch to https:// with the API key, or set SALIENT_EXPOSE_OLLAMA_HTTP=1 in user data to keep the old plain-HTTP listener.
Additional details
Usage instructions
Before you subscribe/launch: GPU instance types need the EC2 quota "Running On-Demand G and VT instances" (quota code L-DB2E81BA) in your launch Region to be at least the instance's vCPU count (g6.xlarge/g5.xlarge/g6e.xlarge = 4 vCPUs; g6.2xlarge/g5.2xlarge/g6e.2xlarge = 8). New AWS accounts often have 0 and the launch fails with "VcpuLimitExceeded". Check or request an increase (usually approved within hours) at https://console.aws.amazon.com/servicequotas/home/services/ec2/quotas/L-DB2E81BA
Launch on g6.xlarge (recommended, lowest cost) or any listed type; g6e.xlarge/g6e.2xlarge (48 GB GPU) give the most headroom for long contexts. Security Group: allow inbound TCP 443 only from the addresses that will call the API; TCP 22 only from your own IP if you need SSH.
Wait about 6 minutes after the instance is running. EC2 console > instance > Actions > Monitor and troubleshoot > Get system log shows "Salient LLM appliance: READY" with the endpoint, API key and TLS fingerprint. Health check: curl -k https://<PUBLIC_IP>/healthz
First request (API key is in the system log, the SSH login banner, or /home/ubuntu/salient-api-key):
curl -k https://<PUBLIC_IP>/api/tags -H "Authorization: Bearer <API_KEY>"
curl -k https://<PUBLIC_IP>/api/generate -H "Authorization: Bearer <API_KEY>" -d '{"model":"orcarouter/Qwen3.8-27B-Uncensored","prompt":"Hello","stream":false}'
OpenAI-compatible: curl -k https://<PUBLIC_IP>/v1/chat/completions -H "Authorization: Bearer <API_KEY>" -H "Content-Type: application/json" -d '{"model":"orcarouter/Qwen3.8-27B-Uncensored","messages":[{"role":"user","content":"Hello"}]}'
TLS is self-signed: use -k or pin the fingerprint from the system log. For a trusted cert set SALIENT_DOMAIN and SALIENT_LETSENCRYPT_EMAIL in user data (needs DNS + TCP 80), or run sudo salient-enable-letsencrypt DOMAIN EMAIL.
Plain HTTP on port 11434 is OFF (Ollama listens on 127.0.0.1 only). For local access use an SSH tunnel: ssh -L 11434:127.0.0.1:11434 ubuntu@<PUBLIC_IP>. Legacy clients can set SALIENT_EXPOSE_OLLAMA_HTTP=1 in user data, then open 11434 to your IP only (no auth on that port).
Troubleshooting: launch error = GPU quota (above). /healthz not ready after 15 minutes = check the system log; contact support@salientengineering.com.
Settings (own API key, SALIENT_ALLOWED_ORIGINS, context length) go in /etc/salient/salient.conf or user data.
Our team is happy to assist with deployment and configuration issues.
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.
A self-hosted, production-ready Qwen 3.8 27B model deployed into your AWS environment with a single click. Because everything runs entirely within your private cloud, your data stays secure, isolated, and fully under your control. Best of all, unlimited tokens.
A self-hosted, production-ready Qwen 3.6 35B model, with 3 billion active parameters deployed into your AWS environment with a single click. Because everything runs entirely within your private cloud, your data stays secure, isolated, and fully under your control. Best of all, unlimited tokens.
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.
Chat, reasoning and image understanding from the open Qwen3.8-27B model, served on one GPU with vLLM. Runs on a SageMaker endpoint in your own AWS account with no network access, so prompts and images are not sent to a third-party API.
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.