Last Updated: August 25, 2025
Deploy from the SaladCloud Portal.

​
Overview

This guide covers deploying GPT-OSS-120B with vLLM on SaladCloud Secure (DataCenter GPUs). GPT-OSS-120B is OpenAI’s largest open-weight model (Apache 2.0 license), designed for advanced reasoning, general-purpose tasks, and large-scale experimentation. It is distributed via Hugging Face without gating, so you can deploy it directly without requiring a token.

​
Key Features

  • Permissive Apache 2.0 license — Build freely without copyleft restrictions or patent risk—ideal for experimentation, customization, and commercial deployment.
  • Configurable reasoning effort — Easily adjust the reasoning effort (low, medium, high) based on your specific use case and latency needs.
  • Full chain-of-thought — Gain complete access to the model’s reasoning process, facilitating easier debugging and increased trust in outputs.
  • Fine-tunable — Fully customize the model to your specific use case through parameter fine-tuning.
  • Agentic capabilities — Use the model’s native capabilities for function calling, web browsing, Python code execution, and structured outputs.
On SaladCloud Secure, this recipe is tuned for 8× H100 GPUs, using tensor parallelism (TP=8) for maximum throughput.

​
Configuration

This recipe comes pre-configured for GPT-OSS-120B. You don’t need to provide model IDs or parallelism settings. Only one option is exposed:
  • GPU Memory Utilization — Fraction of GPU VRAM vLLM may use (default: 0.95). Lower this if you want extra memory headroom for monitoring or sidecar processes.
Everything else is already set for you

​
Example Request

Submit chat completion requests to the /v1/chat/completions endpoint, and receive generated text in response. 
curl https://<YOUR-GATEWAY-URL>/v1/chat/completions \
  -X POST \
  -H 'Content-Type: application/json' \
  -H 'Salad-Api-Key: <YOUR_API_KEY>' \
  -d '{
    "model": "openai/gpt-oss-120b",
    "messages": [
      {"role": "system", "content": "You are a helpful assistant."},
      {"role": "user", "content": "What is deep learning?"}
    ],
    "stream": true,
    "max_tokens": 20
  }'

​
How To Use This Recipe

​
Authentication

If authentication is enabled, requests must include your SaladCloud API key in the Salad-Api-Key header. See Sending Requests.

​
Replica Count

We recommend at least 2 replicas for development and 3–5 replicas for production. Datacenter GPU nodes are still interruptable.

​
Logging

Logs are available in the SaladCloud Portal. You can also connect an external logging provider such as Axiom.

​
Deploy & Wait

When you deploy, SaladCloud will provision nodes, pull the container image, and download the GPT-OSS-120B model weights. First startup may take 15–20 minutes or more. Once replicas show a green checkmark in the Ready column, the service is live.