
How to run GLM-5.3
Bedrock, Vertex AI, or your own GPUs · Checked November 21, 2026
GLM-5.3 has 753B parameters, so it needs about 1,808GB of GPU memory at full precision and about 904GB at FP8. A practical starting point is FP8 on 8x 141GB GPUs.
You have three ways to run it: a managed host that already serves it, your own GPUs in your cloud, or both behind one endpoint. Then you point your agents at it. Here is what each path takes.
GLM-5.3 hardware requirements
| Precision | Memory needed | Smallest setup that fits |
|---|---|---|
| BF16 (full precision) | ~1,808GB | Multiple 8x H200 nodes (about 2) |
| FP8 | ~904GB | 8x 141GB GPUs (one H200 node) |
| 4-bit quantized | ~452GB | 8x 80GB GPUs (one H100 node) |
Our estimates: 753B parameters times bytes per parameter, plus 20% for the KV cache and the serving runtime. Using the full 1M-token context needs more. Check the model card for the precision the weights ship in and any recommended settings.
Where to run GLM-5.3
01Amazon Bedrock
AvailableGLM-5.3 is available on Amazon Bedrock as a managed, serverless model. You call it from your AWS account with IAM, and AWS runs the GPUs. Source
02Google Cloud Vertex AI
Not confirmed as of November 21, 2026We could not confirm GLM-5.3 in Vertex AI Model Garden. To keep it in Google Cloud, serve it yourself on GPUs in your project (GKE or Compute Engine) with the setup below.
03Self-host on your own GPUs
vLLM and SGLang support confirmedDownload zai-org/GLM-5.3 from Hugging Face and serve it with an inference server that exposes an OpenAI-compatible endpoint. vLLM and SGLang support it. Start with FP8 on 8x 141GB GPUs (one H200 node), then load-test with your real prompts before you size for production.
- Keep the endpoint private to your VPC, so prompts and payment data never leave your cloud.
- Read the GLM-5.3 License before you deploy, since custom licenses can carry conditions.
- Run two replicas or keep a hosted model as a fallback, so agents keep working during an upgrade.
Run GLM-5.3 with Runtime
Serving the model is half the job. The other half is giving agents a safe place to use it. Runtime is the AI agent harness for payment and fintech teams, and it is model-neutral: connect GLM-5.3 from Bedrock, Vertex AI, or your own endpoint, and your agents use it next to Claude, GPT, or Gemini.
- In your cloud. Runtime runs in your AWS, GCP, or Azure account or fully self-hosted, next to the model.
- Routed by evals. Evals show which steps GLM-5.3 handles as well as a frontier model, so routine work moves to the cheaper model.
- Fallbacks. If your endpoint is down or slow, agents fall back to another model instead of stopping.
- Guardrails. Agents start read-only, wait for approval before money moves, and every run is recorded.
Harnesses that can drive GLM-5.3 on Runtime:
Running GLM-5.3: FAQ
How much GPU memory does GLM-5.3 need?
+
About 1808GB at BF16, 904GB at FP8, and 452GB at 4 bits. These are our estimates: the weights plus 20% headroom. Long contexts need more memory for the KV cache.
What is the cheapest way to self-host GLM-5.3?
+
Start with FP8 on 8x 141GB GPUs (one H200 node). Quantizing further saves memory but can cost accuracy, so test it on your own tasks first.
Is GLM-5.3 available on Amazon Bedrock?
+
Yes. GLM-5.3 is available on Amazon Bedrock as a managed, serverless model.
Is GLM-5.3 available on Google Vertex AI?
+
As of November 21, 2026, we could not confirm GLM-5.3 in Vertex AI Model Garden. You can still run it in your Google Cloud project on your own GPUs (GKE or Compute Engine) with vLLM and SGLang.
Can I use GLM-5.3 with Runtime?
+
Yes. Runtime is model-neutral. Serve GLM-5.3 from Bedrock, Vertex AI, or your own endpoint, connect it to Runtime, and agents use it through a harness like Claude Code. Your data stays in your cloud, and evals tell you which steps it can take over from a more expensive model.