Is this page helpful?

Get Started with GLM-5.3-Flash#

GLM-5.3-Flash is a vision-language Mixture-of-Experts model from Z.ai with a 1,048,576-token context window. It accepts text and image input, returns reasoning output, and supports tool calling. On the throughput-optimized GB200 NVFP4 profile, the model’s built-in multi-token-prediction (MTP) head serves as the draft model for speculative decoding, so you do not need a separate draft model. The other profiles run without speculative decoding.

GLM-5.3-Flash is served by the SGLang Model-Free NIM, which includes the following for this model:

Fixes for GLM-5.3-Flash image input and for tool_choice="required".
The GLM reasoning parser (glm45) and tool-call parser (glm47), enabled by default.
A throughput-optimized profile for GB200 with the NVFP4 checkpoint, which NIM selects automatically.

The container does not bundle the weights. Set NIM_MODEL_PATH to the FP8, NVFP4, or BF16 checkpoint you want, and the container downloads it at startup, so one image serves all three precisions.

This page covers the deployment settings specific to this model.

Prerequisites#

Before deploying a NIM LLM and VLM container, ensure your environment meets the following requirements:

Hardware Requirements#

The following are the minimum required specifications for supported hardware components:

Requirement	Specification
CPU	AMD64, ARM64
GPU	Refer to the Support Matrix

Software Requirements#

Minimum required versions for supported software components.

Requirement	Specification
Operating System	Ubuntu 22.04 LTS or later recommended
Container Toolkit	1.14.0 or later
CUDA SDK	12.9 or later
GPU Driver	580 or later
Docker	24.0 or later

Operating System#

While other Linux distributions can be compatible with NIM, they have not been officially validated.

We recommend using Ubuntu 22.04 LTS or later for the best experience.

CUDA SDK#

Install CUDA SDK by following the CUDA installation guide for Linux.

GPU Drivers#

Install the NVIDIA GPU drivers by following the NVIDIA Driver Installation Guide.

Docker#

Docker is required to run the containerized NIM services.

Install Docker Engine for your Linux distribution by following the Docker Engine installation guide.
Verify that the Docker daemon is running and that your user can execute docker commands without sudo. Add your user to the docker group if needed:
```
sudo groupadd docker
sudo usermod -aG docker $USER
```
Log out and back in for the group change to take effect.

Container Toolkit#

The NVIDIA Container Toolkit enables Docker containers to access the host GPU.

Install the toolkit by following the NVIDIA Container Toolkit installation guide.
Configure Docker to use the NVIDIA runtime by following the Docker configuration steps.
Restart the Docker daemon after configuration:
```
sudo systemctl restart docker
```

NIM Container Access#

To download and deploy NIM containers, you need one of the following:

A free NVIDIA Developer Program membership.
An NVIDIA AI Enterprise license. To request a free 90-day evaluation license, refer to Ways to Get Started With NVIDIA AI Enterprise and Activate Your NVIDIA AI Enterprise License.

Generate Access Credentials#

NGC#

An NGC Personal API key is not required to access most NVIDIA NIM containers and models hosted on NGC. An NGC API key is only required to download Production Branch (PB) models. This Get Started documentation uses a feature branch model as an example, so creating an NGC Personal API key is optional.

If you want to create an NGC Personal API key, do the following:

Generate the Personal API Key on the Setup API Keys page.
When creating the Personal API key, select at least NGC Catalog from the Services Included list. You can also include additional services if you want to use the same key for other purposes.

Warning

Legacy API keys are not supported by NIM. Always use a Personal API Key.

Hugging Face#

Create a Hugging Face access token with read permissions (or higher, only if your org or private repos require it). This token allows model-free NIM to fetch models directly from Hugging Face.

Verify NVIDIA Runtime Access#

To ensure that your setup is correct, run the following command:

docker run --rm --runtime=nvidia --gpus all ubuntu nvidia-smi

This command should produce output similar to one of the following, where you can confirm CUDA driver version, and available GPUs.

+-----------------------------------------------------------------------------------------+
| NVIDIA-SMI 580.95.05              Driver Version: 580.95.05      CUDA Version: 12.9     |
|-----------------------------------------+------------------------+----------------------+
| GPU  Name                 Persistence-M | Bus-Id          Disp.A | Volatile Uncorr. ECC |
| Fan  Temp   Perf          Pwr:Usage/Cap |           Memory-Usage | GPU-Util  Compute M. |
|                                         |                        |               MIG M. |
|=========================================+========================+======================|
|   0  NVIDIA H100 80GB HBM3          On  |   00000000:1B:00.0 Off |                    0 |
| N/A   36C    P0            112W /  700W |   78489MiB /  81559MiB |      0%      Default |
|                                         |                        |             Disabled |
+-----------------------------------------+------------------------+----------------------+

+-----------------------------------------------------------------------------------------+
| Processes:                                                                              |
|  GPU   GI   CI        PID   Type   Process name                              GPU Memory |
|        ID   ID                                                               Usage      |
|=========================================================================================|
|  No running processes found                                                             |
+-----------------------------------------------------------------------------------------+

Configuration#

NIM LLM and VLM offers several configuration options using environment variables to control caching, logging, and model profiles.

Export a Hugging Face Access Token#

Export a Hugging Face token to download models from Hugging Face.

Export the variable in your shell (temporary), replacing <token-value> with your actual token:
```
export HF_TOKEN="<token-value>"
```

Optional: Persist the variable in your shell profile for future terminals.

If using bash:

echo 'export HF_TOKEN="<token-value>"' >> ~/.bashrc

If using zsh:

echo 'export HF_TOKEN="<token-value>"' >> ~/.zshrc

Verify the variable is set:
```
echo "$HF_TOKEN"
```

Export an NGC API Key#

Note

An NGC API key is only required to download Production Branch (PB) models. You can skip this section if you did not create an NGC API key.

Export the variable in your shell (temporary), replacing <VALUE> with your actual API key:
```
export NGC_API_KEY=<VALUE>
```

Optional: Persist the variable in your shell profile for future terminals.

If using bash:

echo "export NGC_API_KEY=$NGC_API_KEY" >> ~/.bashrc

If using zsh:

echo "export NGC_API_KEY=$NGC_API_KEY" >> ~/.zshrc

Verify the variable is set:
```
echo "$NGC_API_KEY"
```

Model Cache#

NIM downloads model weights to a cache on the host that you mount into the container. Artifacts persist across restarts, so you do not pull the full model on every run.

Local Cache#

An essential variable to configure on your host system is the cache path directory. This directory is mapped from the host machine to container; assets (for example, model weights) are downloaded to this host directory and persist across container restarts. Configuring a local cache is highly recommended, as it avoids re-downloading large model files upon subsequent container restarts. You can name the environment variable containing the path to the local cache whatever you want.

Create the cache directory and export an environment variable:

export LOCAL_NIM_CACHE=~/.cache/nim
mkdir -p $LOCAL_NIM_CACHE
# Optionally add sticky bit to avoid issues writing to the cache if the container is running as a different user
chmod -R a+rwxt $LOCAL_NIM_CACHE

When you start the NIM container, you must map your host machine’s local cache directory ($LOCAL_NIM_CACHE) to the container’s internal cache path (/opt/nim/.cache) using a Docker volume mount, such as -v "$LOCAL_NIM_CACHE:/opt/nim/.cache". This mapping ensures that the large model weights downloaded by the container are saved to your host machine. Because containers are ephemeral, any data stored only inside the container is lost when it stops. By using a volume mount, subsequent container runs detect the existing model files in your local cache and skip the lengthy download process, allowing the NIM to start up faster.

Cache Directory Permissions#

The NIM container runs as a non-root user with GID 0 (root group). The cache directory on your host must be writable by GID 0:

sudo chgrp -R 0 "$LOCAL_NIM_CACHE"
sudo chmod -R g+rwX "$LOCAL_NIM_CACHE"

Run the container with the cache mounted:

docker run --gpus all \
  -v "$LOCAL_NIM_CACHE:/opt/nim/.cache" \
  ...

To run as a custom user (for example, your host user), pass -u <uid>:0:

docker run --gpus all -u $(id -u):0 \
  -v "$LOCAL_NIM_CACHE:/opt/nim/.cache" \
  ...

Important

When using -u <uid>, you must include :0 to set GID 0 (for example, -u $(id -u):0). The container’s writable directories are group-owned by GID 0. Without it, the container will fail with PermissionError when writing to cache, config, or log paths.

Tip

To make this setting permanent across terminal sessions, you can add export LOCAL_NIM_CACHE=~/.cache/nim to your ~/.bashrc or ~/.zshrc profile.

Authenticate#

If you created an NGC API key, authenticate with the NVIDIA Container Registry using the following command:

echo "$NGC_API_KEY" | docker login nvcr.io --username '$oauthtoken' --password-stdin

Use $oauthtoken as the username and NGC_API_KEY as the password. The $oauthtoken username is a special name that indicates that you will authenticate with an API key and not a user name and password.

Choose a Checkpoint#

Set NIM_MODEL_PATH to the checkpoint for the precision you want. For every supported combination of GPU, GPU count, precision, checkpoint, and profile ID, you can find the correct path at GLM-5.3-Flash in the Support Matrix.

Each combination is fixed: the precisions available depend on both the GPU type and the number of GPUs. For example, BF16 needs eight GPUs on B200, H200, and H200 NVL but four on B300, NVFP4 needs two GPUs on B300, and H100 GPUs need eight GPUs for FP8 and NVFP4. NIM selects the profile from the GPUs it detects and the precision of the checkpoint in NIM_MODEL_PATH. To override the automatic selection, refer to Model Profiles and Selection.

List the Applicable Profiles#

To list the profiles available on your system, run list-model-profiles with the same NIM_MODEL_PATH that you used in the previous section. Because the container does not bundle the weights, it needs the checkpoint location to build its profile list; without NIM_MODEL_PATH, the command fails.

docker run --rm --gpus all \
  -e NGC_API_KEY \
  -e HF_TOKEN \
  -e NIM_MODEL_PATH=hf://nvidia/GLM-5.3-Flash-NVFP4 \
  -v "$LOCAL_NIM_CACHE:/opt/nim/.cache" \
  -u $(id -u):0 \
  nvcr.io/nim/nvidia/sglang-model-free-nim:2.1.4-variant-glm-5.3-flash list-model-profiles

To check which profile a running NIM selected, query /v1/metadata and read profile_id and profile_name.

On H100 and H200 GPUs, NVFP4 checkpoints run with Marlin weight-only kernels: the weights stay in 4-bit and the math runs in BF16. This uses less memory than FP8 but is slower than FP8 on the same GPUs. This container does not need NIM_ALLOW_NVFP4_EMULATION=1 for the supported H100 and H200 NVFP4 configurations: their profiles are validated for Marlin and enabled by default. The list-model-profiles output can show warnings that other NVFP4 profiles are not tagged for emulation. These warnings are expected, and NIM still selects the validated profile.

The following GPUs are not supported:

RTX PRO 6000 Blackwell and RTX PRO 4500 Blackwell: SGLang 0.5.20 has no sparse-attention (DSA) backend for this GPU architecture.
A100 and L40S: the sparse-attention kernels need Hopper (SM90) or newer.
GH200: the Arm64 SGLang kernel package does not include the FlashAttention 3 kernels that Hopper sparse-attention decoding uses.
A10G and GB10: the model does not fit in GPU memory on one node.

Every profile runs on a single node. Multi-node deployment is not supported.

Throughput-Optimized GB200 NVFP4 Profile#

On four GB200 GPUs with the NVFP4 checkpoint, NIM selects a throughput-optimized profile. It uses MTP speculative decoding and a hierarchical KV cache that keeps 100 GB of KV cache per GPU in host memory. The host needs about 400 GB of free RAM for the four GPUs, and the container needs --shm-size=64g. To turn off speculative decoding, set NIM_SPECDEC_ENABLE=0.

Run NIM#

To run the container and download the model, do the following:

Start the container. Set NIM_MODEL_PATH to the checkpoint listed for your GPU and GPU count in GLM-5.3-Flash. The following example runs the throughput-optimized profile on four GB200 GPUs:
```
docker run -it --rm \
  --gpus all \
  --shm-size=64g \
  -e NGC_API_KEY \
  -e HF_TOKEN \
  -e NIM_MODEL_PATH=hf://nvidia/GLM-5.3-Flash-NVFP4 \
  -e NIM_SERVED_MODEL_NAME=zai-org/glm-5.3-flash \
  -v "$LOCAL_NIM_CACHE:/opt/nim/.cache" \
  -u $(id -u):0 \
  -p 8000:8000 \
  nvcr.io/nim/nvidia/sglang-model-free-nim:2.1.4-variant-glm-5.3-flash
```
The example uses --gpus all. On a host with more GPUs than the profile uses, such as an eight-GPU H200 host running a four-GPU profile, pass only the GPUs that you want NIM to use, for example --gpus '"device=0,1,2,3"', and set NIM_MODEL_PATH to the checkpoint for that GPU type from GLM-5.3-Flash.

NIM_SERVED_MODEL_NAME sets the name that clients send in the model field. Without it, NIM derives the name from the checkpoint in NIM_MODEL_PATH, so the name changes with the precision you select. Refer to Environment Variables for details.

The first start downloads about 205 GB (NVFP4), 330 GB (FP8), or 645 GB (BF16) of weights. To serve a checkpoint you already downloaded, refer to Model Download.
In another terminal, wait until the server is ready. The Application startup complete log line can appear one to two minutes before the server accepts requests, so poll the readiness endpoint until it returns HTTP 200:
```
until [ "$(curl -s -o /dev/null -w '%{http_code}' http://localhost:8000/v1/health/ready)" = "200" ]; do sleep 30; done
curl -s http://localhost:8000/v1/health/ready
```
A ready server returns {"object":"health.response","message":"ready","status":"ready"}.

Interact with the API#

This NIM runs SGLang, not vLLM (API Reference describes the vLLM backend). On this NIM, the following endpoints are available:

/v1/chat/completions
/v1/completions
/v1/responses
/v1/messages
/v1/messages/count_tokens
/v1/models
/tokenize
/detokenize

Note

The following vLLM-only endpoints return HTTP 404 and are not supported:

/inference/v1/generate
/generative_scoring
/v1/chat/completions/render
/v1/completions/render

Response storage is off, so previous_response_id and GET /v1/responses/{response_id} are not supported.

Send a Chat Completion Request#

Send the following request:

curl -s http://localhost:8000/v1/chat/completions \
  -H "Content-Type: application/json" \
  -d '{
    "model": "zai-org/glm-5.3-flash",
    "messages": [{"role": "user", "content": "How many prime numbers are there below 30?"}],
    "max_tokens": 8192
  }'

The response carries the reasoning in message.reasoning_content and the answer in message.content. Reasoning tokens count toward max_tokens. If finish_reason is length and content is empty, the model used the whole budget on reasoning: increase max_tokens or lower the reasoning effort (refer to Control Reasoning).

Control Reasoning#

Reasoning cannot be turned off. The default reasoning effort is max. For shorter reasoning and faster answers, set reasoning_effort to low or high.

Send the following request:

curl -s http://localhost:8000/v1/chat/completions \
  -H "Content-Type: application/json" \
  -d '{
    "model": "zai-org/glm-5.3-flash",
    "messages": [{"role": "user", "content": "What is GPU computing?"}],
    "reasoning_effort": "low",
    "max_tokens": 8192
  }'

You can also set the effort as "chat_template_kwargs": {"reasoning_effort": "low"}. The model supports low, high, and max. The API also accepts none, minimal, medium, and xhigh, but the model runs them at max. Setting "chat_template_kwargs": {"enable_thinking": false} does not turn reasoning off.

stop sequences also match text in the reasoning. If a stop sequence appears while the model is reasoning, the response ends before the answer and content is empty.

When you stream the response, all delta.reasoning_content chunks arrive before the first delta.content chunk.

Enable Tool Calling#

Send the following request:

curl -s http://localhost:8000/v1/chat/completions \
  -H "Content-Type: application/json" \
  -d '{
    "model": "zai-org/glm-5.3-flash",
    "messages": [{"role": "user", "content": "What is the weather in Paris?"}],
    "tools": [{"type": "function", "function": {
      "name": "get_weather",
      "description": "Get the current weather for a city",
      "parameters": {"type": "object", "properties": {"city": {"type": "string"}}, "required": ["city"]}
    }}],
    "tool_choice": "auto",
    "max_tokens": 8192
  }'

tool_choice accepts auto, required, none, or a named function, such as {"type": "function", "function": {"name": "get_weather"}}.

Send an Image#

Send the following request:

curl -s http://localhost:8000/v1/chat/completions \
  -H "Content-Type: application/json" \
  -d '{
    "model": "zai-org/glm-5.3-flash",
    "messages": [{"role": "user", "content": [
      {"type": "text", "text": "Describe this image."},
      {"type": "image_url", "image_url": {"url": "https://assets.ngc.nvidia.com/products/api-catalog/phi-3-5-vision/example1b.jpg"}}
    ]}],
    "max_tokens": 8192
  }'

Video input is not validated on this NIM. Use text and image input.