Multi-GPU
Oracle Cloud runbook · model rank #18

Stand up GLM-5
on Oracle Cloud.

Fits one high-memory node, but needs tensor/expert parallel configuration. Bare-metal GPU shapes and OCI-native imported-model or Kubernetes deployments.

Open Oracle Cloud ↗ Read the runbook
Multi-GPU: Fits one high-memory node, but needs tensor/expert parallel configuration.
Provider walkthrough

Oracle Cloud setup, in the order that matters

OCI Generative AI imported models when supported; OKE vLLM otherwise. Use the provider console for infrastructure and the pinned runtime block for the model server.

  1. 01

    Choose imported model or OKE

    Use OKE with the vLLM Production Stack because this exact family is not in Oracle's explicitly tested imported-model list.

  2. 02

    Prepare model access and capacity

    Use Hugging Face or OCI Object Storage for zai-org/GLM-5, reserve OCI GPU shape with at least 8x H100/A100 80GB-class accelerators, and allocate at least 1000GB for weights and engine overhead.

  3. 03

    Configure the engine

    Let Open Model Engine select vLLM/SGLang for a compatible import, or apply the runtime block below in OKE with an initial 32,768-token limit.

  4. 04

    Test and clean up

    Verify the chat endpoint, inspect GPU memory and logs, then delete unused endpoints, node pools, and boot volumes to stop charges.

Runtime seedvLLM · v0.19.0
docker run --rm --gpus all --ipc=host \
  -p 8000:8000 \
  -e HUGGING_FACE_HUB_TOKEN="$HF_TOKEN" \
  -v "$PWD/.hf-cache:/root/.cache/huggingface" \
  vllm/vllm-openai:v0.19.0 \
  --model zai-org/GLM-5 \
  --tensor-parallel-size 8 \
  --max-model-len 32768 \
  --served-model-name glm-5 \
  --enable-auto-tool-choice \
  --tool-call-parser glm47 \
  --reasoning-parser glm45

This is a reproducible starting block. Cluster paths still need the model author's distributed recipe and the provider's orchestration layer.

Smoke testOpenAI-compatible request
curl http://127.0.0.1:8000/v1/chat/completions \
  -H "Content-Type: application/json" \
  -d '{
    "model": "glm-5",
    "messages": [{"role": "user", "content": "Reply with: deployment healthy"}],
    "max_tokens": 32
  }'

Run this only after logs show the model loaded. A successful HTTP response is not a latency or quality benchmark.

Before you launch

Quota, license, and storage

  • Confirm OCI GPU shape with at least 8x H100/A100 80GB-class accelerators is actually available in the selected region.
  • Read the MIT terms and accept any gated-model conditions.
  • Budget at least 1000GB for weights, cache, and container layers.
  • Keep HF_TOKEN in the provider secret store—not in scripts or templates.
Before real traffic

Memory, format, and shutdown

  • Record idle/free VRAM after the model loads and after a representative prompt.
  • Validate the official chat template, reasoning parser, and tool-call parser.
  • Add authentication and TLS in front of port 8000.
  • Verify the provider's stop/delete action actually ends compute billing.
Agent-assisted setup

Hand this runbook to Claude Code or Codex

Open a terminal in the repository where you want the deployment files, start claude or codex, then paste this prompt. It asks the agent to verify sources and stop before it creates billable infrastructure.

Claude Code quickstart ↗ Codex docs ↗
Deployment prompt 
Deploy GLM-5 (zai-org/GLM-5) on Oracle Cloud. Use this guide as the starting context: https://www.getflops.ai/models/glm-5/oracle. Use a planning floor of 8x 96GB GPUs, 1000GB storage, vllm v0.19.0, and an initial context limit of 32768 tokens. Open every linked primary source and flag any mismatch instead of guessing. Create a deployment folder containing README.md, .env.example with no secrets, a pinned start script or infrastructure manifest, and smoke-test.sh. Make the endpoint OpenAI-compatible where the runtime supports it. Run local/static validation, estimate the billable resources, and stop before provisioning paid infrastructure until I approve.
Guardrails included No secrets in files · verify primary docs · approval before spend
Research trail

The two sources to check first

Z.ai model card Model size, context, license, chat format, and engine notes ↗ OCI imported-model compatibility docs OCI Generative AI imported models when supported; OKE vLLM otherwise ↗

Checked 2026-07-27. This page separates sourced facts from the planning baseline; verify current runtime support and capacity before spending.

Same model, another cloud

Compare GLM-5 elsewhere

RunPod Multi-GPU 8x H100/A100 80GB-class GPUs Vast.ai Multi-GPU 8x H100/A100 80GB-class GPUs Lambda Multi-GPU 8x H100/A100 80GB-class GPUs AWS Multi-GPU SageMaker/HyperPod GPU capacity with at least 8x H100/A100 80GB-class accelerators Google Cloud Multi-GPU Vertex AI GPU deployment with at least 8x H100/A100 80GB-class accelerators Azure Multi-GPU Azure ML GPU deployment with at least 8x H100/A100 80GB-class accelerators