> ## Documentation Index
> Fetch the complete documentation index at: https://docs.sglang.io/llms.txt
> Use this file to discover all available pages before exploring further.

# DeepSeek-V3.2

export const DeepSeekV32Deployment = () => {
  const options = {
    hardware: {
      name: 'hardware',
      title: 'Hardware Platform',
      items: [{
        id: 'h200',
        label: 'H200',
        default: true
      }, {
        id: 'b200',
        label: 'B200',
        default: false
      }, {
        id: 'mi300x',
        label: 'MI300X',
        default: false
      }, {
        id: 'mi355x',
        label: 'MI355X',
        default: false
      }]
    },
    modelname: {
      name: 'modelname',
      title: 'Model Name',
      getDynamicItems: values => {
        const hw = values.hardware;
        const isB200 = hw === 'b200';
        const isAMD = hw === 'mi300x' || hw === 'mi355x';
        return [{
          id: 'v32',
          label: 'DeepSeek-V3.2',
          default: !isB200 && !isAMD
        }, {
          id: 'v32speciale',
          label: 'DeepSeek-V3.2-Speciale',
          default: false
        }, {
          id: 'v32exp',
          label: 'DeepSeek-V3.2-Exp',
          default: false
        }, {
          id: 'v32nvfp4',
          label: 'DeepSeek-V3.2-NVFP4',
          default: isB200,
          disabled: !isB200,
          disabledReason: 'NVFP4 requires B200 (Blackwell)'
        }, {
          id: 'v32mxfp4',
          label: 'DeepSeek-V3.2-MXFP4',
          default: isAMD,
          disabled: !isAMD,
          disabledReason: 'MXFP4 requires AMD MI300X/MI355X'
        }];
      }
    },
    strategy: {
      name: 'strategy',
      title: 'Deployment Strategy',
      type: 'checkbox',
      condition: values => values.modelname !== 'v32nvfp4' && values.modelname !== 'v32mxfp4',
      items: [{
        id: 'tp',
        label: 'TP',
        default: true,
        required: true
      }, {
        id: 'dp',
        label: 'DP attention',
        default: false
      }, {
        id: 'ep',
        label: 'EP',
        default: false
      }, {
        id: 'mtp',
        label: 'Multi-token Prediction',
        default: false
      }]
    },
    reasoningParser: {
      name: 'reasoningParser',
      title: 'Reasoning Parser',
      condition: values => values.modelname !== 'v32nvfp4' && values.modelname !== 'v32mxfp4',
      items: [{
        id: 'disabled',
        label: 'Disabled',
        default: true
      }, {
        id: 'enabled',
        label: 'Enabled',
        default: false
      }]
    },
    toolcall: {
      name: 'toolcall',
      title: 'Tool Call Parser',
      condition: values => values.modelname !== 'v32nvfp4' && values.modelname !== 'v32mxfp4' && values.modelname !== 'v32speciale',
      items: [{
        id: 'disabled',
        label: 'Disabled',
        default: true
      }, {
        id: 'enabled',
        label: 'Enabled',
        default: false
      }]
    }
  };
  const resolveItems = (option, vals) => {
    if (typeof option.getDynamicItems === 'function') return option.getDynamicItems(vals);
    return option.items;
  };
  const getInitialState = () => {
    const initialState = {};
    for (const [key, option] of Object.entries(options)) {
      if (option.type === 'checkbox') {
        const items = resolveItems(option, initialState);
        initialState[key] = items.filter(i => i.default).map(i => i.id);
      } else {
        const items = resolveItems(option, initialState);
        const def = items.find(i => i.default && !i.disabled) || items.find(i => !i.disabled) || items[0];
        initialState[key] = def.id;
      }
    }
    return initialState;
  };
  const [values, setValues] = useState(getInitialState);
  const [isDark, setIsDark] = useState(false);
  useEffect(() => {
    const checkDarkMode = () => {
      const html = document.documentElement;
      const isDarkMode = html.classList.contains('dark') || html.getAttribute('data-theme') === 'dark' || html.style.colorScheme === 'dark';
      setIsDark(isDarkMode);
    };
    checkDarkMode();
    const observer = new MutationObserver(checkDarkMode);
    observer.observe(document.documentElement, {
      attributes: true,
      attributeFilter: ['class', 'data-theme', 'style']
    });
    return () => observer.disconnect();
  }, []);
  useEffect(() => {
    setValues(prev => {
      const next = {
        ...prev
      };
      for (const [key, option] of Object.entries(options)) {
        if (typeof option.getDynamicItems !== 'function') continue;
        const items = option.getDynamicItems(next);
        const current = items.find(i => i.id === next[key]);
        if (!current || current.disabled) {
          const fallback = items.find(i => i.default && !i.disabled) || items.find(i => !i.disabled);
          if (fallback) next[key] = fallback.id;
        }
      }
      return next;
    });
  }, [values.hardware]);
  const handleRadioChange = (optionName, value) => {
    setValues(prev => ({
      ...prev,
      [optionName]: value
    }));
  };
  const handleCheckboxChange = (optionName, itemId, isChecked) => {
    setValues(prev => {
      const currentValues = prev[optionName] || [];
      if (isChecked) {
        return {
          ...prev,
          [optionName]: [...currentValues, itemId]
        };
      } else {
        return {
          ...prev,
          [optionName]: currentValues.filter(id => id !== itemId)
        };
      }
    });
  };
  const generateCommand = () => {
    const {hardware, modelname, strategy, reasoningParser, toolcall} = values;
    const isNvfp4 = modelname === 'v32nvfp4';
    const isMxfp4 = modelname === 'v32mxfp4';
    const isAMD = hardware === 'mi300x' || hardware === 'mi355x';
    if (isNvfp4 && hardware !== 'b200') {
      return `# Error: DeepSeek-V3.2-NVFP4 requires NVIDIA B200 (Blackwell) hardware\n# Please select "B200" for Hardware Platform or choose a different model`;
    }
    if (isMxfp4 && !isAMD) {
      return `# Error: DeepSeek-V3.2-MXFP4 requires AMD MI300X/MI355X hardware\n# Please select "MI300X" or "MI355X" for Hardware Platform or choose a different model`;
    }
    if (modelname === 'v32speciale' && toolcall === 'enabled') {
      return `# Error: DeepSeek-V3.2-Speciale doesn't support tool calling\n# Please select "Disabled" for Tool Call Parser or choose a different model`;
    }
    const modelMap = {
      'v32': 'DeepSeek-V3.2',
      'v32exp': 'DeepSeek-V3.2-Exp',
      'v32speciale': 'DeepSeek-V3.2-Speciale',
      'v32nvfp4': 'DeepSeek-V3.2-NVFP4',
      'v32mxfp4': 'DeepSeek-V3.2-mxfp4'
    };
    let modelFamily;
    if (isNvfp4) modelFamily = 'nvidia'; else if (isMxfp4) modelFamily = 'amd'; else modelFamily = 'deepseek-ai';
    const modelName = `${modelFamily}/${modelMap[modelname]}`;
    if (isNvfp4) {
      let cmd = 'sglang serve \\\n';
      cmd += `  --model ${modelName}`;
      cmd += ' \\\n  --tp 4';
      cmd += ' \\\n  --quantization modelopt_fp4';
      cmd += ' \\\n  --moe-runner-backend flashinfer_trtllm';
      return cmd;
    }
    if (isMxfp4) {
      let cmd = 'sglang serve \\\n';
      cmd += `  --model ${modelName}`;
      cmd += ' \\\n  --tp 8';
      cmd += ' \\\n  --trust-remote-code';
      return cmd;
    }
    let cmd = 'sglang serve \\\n';
    cmd += `  --model ${modelName}`;
    if (isAMD) {
      cmd += ' \\\n  --trust-remote-code';
      cmd += ' \\\n  --nsa-prefill-backend tilelang';
      cmd += ' \\\n  --nsa-decode-backend tilelang';
      cmd += ' \\\n  --cuda-graph-max-bs 64';
    }
    const strategyArray = Array.isArray(strategy) ? strategy : [];
    const tpSize = 8;
    const dpSize = 8;
    const epSize = 8;
    cmd += ` \\\n  --tp ${tpSize}`;
    if (strategyArray.includes('dp')) {
      cmd += ` \\\n  --dp ${dpSize} \\\n  --enable-dp-attention`;
    }
    if (strategyArray.includes('ep')) {
      cmd += ` \\\n  --ep ${epSize}`;
    }
    if (strategyArray.includes('mtp')) {
      cmd += ' \\\n  --speculative-algorithm EAGLE';
      cmd += ' \\\n  --speculative-num-steps 3';
      cmd += ' \\\n  --speculative-eagle-topk 1';
      cmd += ' \\\n  --speculative-num-draft-tokens 4';
    }
    if (toolcall === 'enabled' && modelname !== 'v32speciale') {
      if (modelname === 'v32exp') {
        cmd += ' \\\n  --tool-call-parser deepseekv31';
      } else if (modelname === 'v32') {
        cmd += ' \\\n  --tool-call-parser deepseekv32';
      }
    }
    if (reasoningParser === 'enabled') {
      cmd += ' \\\n  --reasoning-parser deepseek-v3';
    }
    if (toolcall === 'enabled' && modelname === 'v32exp') {
      cmd += ' \\\n  --chat-template ./examples/chat_template/tool_chat_template_deepseekv32.jinja';
    }
    return cmd;
  };
  const containerStyle = {
    maxWidth: '900px',
    margin: '0 auto',
    display: 'flex',
    flexDirection: 'column',
    gap: '4px'
  };
  const cardStyle = {
    padding: '8px 12px',
    border: `1px solid ${isDark ? '#374151' : '#e5e7eb'}`,
    borderLeft: `3px solid ${isDark ? '#E85D4D' : '#D45D44'}`,
    borderRadius: '4px',
    display: 'flex',
    alignItems: 'center',
    gap: '12px',
    background: isDark ? '#1f2937' : '#fff'
  };
  const titleStyle = {
    fontSize: '13px',
    fontWeight: '600',
    minWidth: '140px',
    flexShrink: 0,
    color: isDark ? '#e5e7eb' : 'inherit'
  };
  const itemsStyle = {
    display: 'flex',
    rowGap: '2px',
    columnGap: '6px',
    flexWrap: 'wrap',
    alignItems: 'center',
    flex: 1
  };
  const labelBaseStyle = {
    padding: '4px 10px',
    border: `1px solid ${isDark ? '#9ca3af' : '#d1d5db'}`,
    borderRadius: '3px',
    cursor: 'pointer',
    display: 'inline-flex',
    flexDirection: 'column',
    alignItems: 'center',
    justifyContent: 'center',
    fontWeight: '500',
    fontSize: '13px',
    transition: 'all 0.2s',
    userSelect: 'none',
    minWidth: '45px',
    textAlign: 'center',
    flex: 1,
    background: isDark ? '#374151' : '#fff',
    color: isDark ? '#e5e7eb' : 'inherit'
  };
  const checkedStyle = {
    background: '#D45D44',
    color: 'white',
    borderColor: '#D45D44'
  };
  const disabledStyle = {
    cursor: 'not-allowed',
    opacity: 0.4
  };
  const subtitleStyle = {
    display: 'block',
    fontSize: '9px',
    marginTop: '1px',
    lineHeight: '1.1',
    opacity: 0.7
  };
  const commandDisplayStyle = {
    flex: 1,
    padding: '12px 16px',
    background: isDark ? '#111827' : '#f5f5f5',
    borderRadius: '6px',
    fontFamily: "'Menlo', 'Monaco', 'Courier New', monospace",
    fontSize: '12px',
    lineHeight: '1.5',
    color: isDark ? '#e5e7eb' : '#374151',
    whiteSpace: 'pre-wrap',
    overflowX: 'auto',
    margin: 0,
    border: `1px solid ${isDark ? '#374151' : '#e5e7eb'}`
  };
  return <div style={containerStyle} className="not-prose">
      {Object.entries(options).map(([key, option]) => {
    if (typeof option.condition === 'function' && !option.condition(values)) return null;
    const items = resolveItems(option, values);
    return <div key={key} style={cardStyle}>
            <div style={titleStyle}>{option.title}</div>
            <div style={itemsStyle}>
              {option.type === 'checkbox' ? items.map(item => {
      const isChecked = (values[option.name] || []).includes(item.id);
      const isDisabled = item.required || !!item.disabled;
      return <label key={item.id} style={{
        ...labelBaseStyle,
        ...isChecked ? checkedStyle : {},
        ...isDisabled ? {
          ...disabledStyle,
          ...item.required ? {} : {}
        } : {}
      }} title={item.disabledReason || ''}>
                      <input type="checkbox" checked={isChecked} disabled={isDisabled} onChange={e => handleCheckboxChange(option.name, item.id, e.target.checked)} style={{
        display: 'none'
      }} />
                      {item.label}
                      {item.subtitle && <small style={{
        ...subtitleStyle,
        color: isChecked ? 'rgba(255,255,255,0.85)' : 'inherit'
      }}>{item.subtitle}</small>}
                    </label>;
    }) : items.map(item => {
      const isChecked = values[option.name] === item.id;
      const isDisabled = !!item.disabled;
      return <label key={item.id} style={{
        ...labelBaseStyle,
        ...isChecked ? checkedStyle : {},
        ...isDisabled ? disabledStyle : {}
      }} title={item.disabledReason || ''}>
                      <input type="radio" name={option.name} value={item.id} checked={isChecked} disabled={isDisabled} onChange={() => !isDisabled && handleRadioChange(option.name, item.id)} style={{
        display: 'none'
      }} />
                      {item.label}
                      {item.subtitle && <small style={{
        ...subtitleStyle,
        color: isChecked ? 'rgba(255,255,255,0.85)' : 'inherit'
      }}>{item.subtitle}</small>}
                    </label>;
    })}
            </div>
          </div>;
  })}
      <div style={cardStyle}>
        <div style={titleStyle}>Run this Command:</div>
        <pre style={commandDisplayStyle}>{generateCommand()}</pre>
      </div>
    </div>;
};

## 1. Model Introduction

The DeepSeek-V3.2 series includes three model variants, each optimized for different use cases:

**[DeepSeek-V3.2-Exp](https://huggingface.co/deepseek-ai/DeepSeek-V3.2-Exp)** is an upgraded version of DeepSeek-V3.1-Terminus, introducing the DeepSeek Sparse Attention (DSA) mechanism through continued training. DSA is a fine-grained sparse attention mechanism powered by a lightning indexer, enabling DeepSeek-V3.2-Exp to achieve significant efficiency improvements in long-context scenarios. Recommended for general conversations, long-context processing, and efficient inference.

**[DeepSeek-V3.2](https://huggingface.co/deepseek-ai/DeepSeek-V3.2)** is the standard version suitable for general tasks and conversational scenarios. For local deployment, we recommend setting the sampling parameters to temperature = 1.0, top\_p = 0.95. Recommended for standard conversations and general tasks.

**[DeepSeek-V3.2-Speciale](https://huggingface.co/deepseek-ai/DeepSeek-V3.2-Speciale)** is a special variant designed exclusively for deep reasoning tasks. This model is specifically optimized for scenarios requiring complex logical reasoning and deep thinking. However this model does not support tool calls (see below). For local deployment, we recommend setting the sampling parameters to temperature = 1.0, top\_p = 0.95. Recommended for deep reasoning tasks, complex logical problems, and mathematical reasoning.

**[DeepSeek-V3.2-NVFP4](https://huggingface.co/nvidia/DeepSeek-V3.2-NVFP4)** is an NVIDIA-optimized NVFP4-quantized variant of DeepSeek-V3.2 for Blackwell devices. It uses ModelOpt FP4 quantization with a choice of MoE runner backends (`flashinfer_trtllm` (recommended), `flashinfer_cutlass`, or `flashinfer_cutedsl`), enabling efficient deployment with lower tensor parallelism (TP=4). It supports the same features as DeepSeek-V3.2 including tool calling, reasoning, and speculative decoding (MTP).

**[DeepSeek-V3.2-MXFP4](https://huggingface.co/amd/DeepSeek-V3.2-mxfp4)** is an OCP-MXFP4 optimized variant for DeepSeek-V3.2 for AMD MI300X/MI355X devices. It uses OCP MXFP4 quantization with a triton mxfp4 backend (the same backend for gptoss-120B), enabling efficient deployment with lower tensor parallelism (TP=8) in a single node. It includes the same features as DeepSeek-V3.2 including tool calling, reasoning, fp8-kv, CP, TP and speculative decoding MTP.

## 2. SGLang Installation

SGLang offers multiple installation methods. You can choose the most suitable installation method based on your hardware platform and requirements.

Please refer to the [official SGLang installation guide](../../../docs/get-started/install) for installation instructions.

## 3. Model Deployment

This section provides a progressive guide from quick deployment to performance optimization, suitable for users at different levels.

### 3.1 Basic Configuration

**Interactive Command Generator**: Use the configuration selector below to automatically generate the appropriate deployment command for your hardware platform, model variant, deployment strategy, and thinking capabilities. SGLang supports serving DeepSeek V3.2 on NVIDIA H200, B200, and AMD MI300X/MI355X GPUs.

<DeepSeekV32Deployment />

### 3.2 Configuration Tips

For more detailed configuration tips, please refer to [DeepSeek-V3.2 Usage](../../../docs/basic_usage/deepseek_v32).

## 4. Model Invocation

### 4.1 Basic Usage

For basic API usage and request examples, please refer to:

* [Basic API Usage](../../../docs/basic_usage/send_request)

### 4.2 Advanced Usage

#### 4.2.1 Reasoning Parser

DeepSeek-V3.2 supports reasoning mode. Enable the reasoning parser during deployment to separate the thinking and content sections:

```shell Command theme={null}
sglang serve \
  --model deepseek-ai/DeepSeek-V3.2-Exp \
  --reasoning-parser deepseek-v3 \
  --tp 8 \
  --host 0.0.0.0 \
  --port 30000
```

**Streaming with Thinking Process:**

```python Example theme={null}
from openai import OpenAI

client = OpenAI(
    base_url="http://localhost:30000/v1",
    api_key="EMPTY"
)

# Enable streaming to see the thinking process in real-time
response = client.chat.completions.create(
    model="deepseek-ai/DeepSeek-V3.2-Exp",
    messages=[
        {"role": "user", "content": "Solve this problem step by step: What is 15% of 240?"}
    ],
    temperature=0.7,
    max_tokens=2048,
    extra_body = {"chat_template_kwargs": {"thinking": True}},
    stream=True
)

# Process the stream
has_thinking = False
has_answer = False
thinking_started = False

for chunk in response:
    if chunk.choices and len(chunk.choices) > 0:
        delta = chunk.choices[0].delta

        # Print thinking process
        if hasattr(delta, 'reasoning_content') and delta.reasoning_content:
            if not thinking_started:
                print("=============== Thinking =================", flush=True)
                thinking_started = True
            has_thinking = True
            print(delta.reasoning_content, end="", flush=True)

        # Print answer content
        if delta.content:
            # Close thinking section and add content header
            if has_thinking and not has_answer:
                print("\n=============== Content =================", flush=True)
                has_answer = True
            print(delta.content, end="", flush=True)

print()
```

**Output Example:**

```text Output theme={null}
=============== Thinking =================
To solve this problem, I need to calculate 15% of 240.
Step 1: Convert 15% to decimal: 15% = 0.15
Step 2: Multiply 240 by 0.15
Step 3: 240 × 0.15 = 36
=============== Content =================

The answer is 36. To find 15% of 240, we multiply 240 by 0.15, which equals 36.
```

**Note:** The reasoning parser captures the model's step-by-step thinking process, allowing you to see how the model arrives at its conclusions.

#### 4.2.2 Tool Calling

DeepSeek-V3.2 and DeepSeek-V3.2-Exp support tool calling capabilities. But they use different parameters. Enable the tool call parser:

**Note:** DeepSeek-V3.2-Speciale does **NOT** support tool calling. It is designed exclusively for deep reasoning tasks.

**Deployment Command:**

For DeepSeek-V3.2-Exp:

```shell Command theme={null}
sglang serve \
  --model deepseek-ai/DeepSeek-V3.2-Exp \
  --tool-call-parser deepseekv31 \
  --reasoning-parser deepseek-v3 \
  --chat-template ./examples/chat_template/tool_chat_template_deepseekv32.jinja \
  --tp 8 \
  --host 0.0.0.0 \
  --port 30000
```

For DeepSeek-V3.2, use `--tool-call-parser deepseekv32` and remove `--chat-template`.

**Python Example (with Thinking Process):**

```python Example theme={null}
from openai import OpenAI

client = OpenAI(
    base_url="http://localhost:30000/v1",
    api_key="EMPTY"
)

# Define available tools
tools = [
    {
        "type": "function",
        "function": {
            "name": "get_weather",
            "description": "Get the current weather for a location",
            "parameters": {
                "type": "object",
                "properties": {
                    "location": {
                        "type": "string",
                        "description": "The city name"
                    },
                    "unit": {
                        "type": "string",
                        "enum": ["celsius", "fahrenheit"],
                        "description": "Temperature unit"
                    }
                },
                "required": ["location"]
            }
        }
    }
]

# Make request with streaming to see thinking process
response = client.chat.completions.create(
    model="deepseek-ai/DeepSeek-V3.2-Exp",
    messages=[
        {"role": "user", "content": "What's the weather in Beijing?"}
    ],
    tools=tools,
    extra_body = {"chat_template_kwargs": {"thinking": True}},
    temperature=0.7,
    stream=True
)

# Process streaming response
thinking_started = False
has_thinking = False
tool_calls_accumulator = {}

for chunk in response:
    if chunk.choices and len(chunk.choices) > 0:
        delta = chunk.choices[0].delta

        # Print thinking process
        if hasattr(delta, 'reasoning_content') and delta.reasoning_content:
            if not thinking_started:
                print("=============== Thinking =================", flush=True)
                thinking_started = True
            has_thinking = True
            print(delta.reasoning_content, end="", flush=True)

        # Accumulate tool calls
        if hasattr(delta, 'tool_calls') and delta.tool_calls:
            # Close thinking section if needed
            if has_thinking and thinking_started:
                print("\n=============== Content =================\n", flush=True)
                thinking_started = False

            for tool_call in delta.tool_calls:
                index = tool_call.index
                if index not in tool_calls_accumulator:
                    tool_calls_accumulator[index] = {
                        'name': None,
                        'arguments': ''
                    }

                if tool_call.function:
                    if tool_call.function.name:
                        tool_calls_accumulator[index]['name'] = tool_call.function.name
                    if tool_call.function.arguments:
                        tool_calls_accumulator[index]['arguments'] += tool_call.function.arguments

        # Print content
        if delta.content:
            print(delta.content, end="", flush=True)

# Print accumulated tool calls
for index, tool_call in sorted(tool_calls_accumulator.items()):
    print(f"Tool Call: {tool_call['name']}")
    print(f"   Arguments: {tool_call['arguments']}")

print()
```

**Output Example:**

```text Output theme={null}
=============== Thinking =================
The user is asking about the weather in Beijing. I need to use the get_weather function to retrieve this information.
I should call the function with location="Beijing".
=============== Content =================

Tool Call: get_weather
   Arguments: {"location": "Beijing", "unit": "celsius"}
```

**Note:**

* The reasoning parser shows how the model decides to use a tool
* Tool calls are clearly marked with the function name and arguments
* You can then execute the function and send the result back to continue the conversation

**Handling Tool Call Results:**

```python Example theme={null}
# After getting the tool call, execute the function
def get_weather(location, unit="celsius"):
    # Your actual weather API call here
    return f"The weather in {location} is 22°{unit[0].upper()} and sunny."

# Send tool result back to the model
messages = [
    {"role": "user", "content": "What's the weather in Beijing?"},
    {
        "role": "assistant",
        "content": None,
        "tool_calls": [{
            "id": "call_123",
            "type": "function",
            "function": {
                "name": "get_weather",
                "arguments": '{"location": "Beijing", "unit": "celsius"}'
            }
        }]
    },
    {
        "role": "tool",
        "tool_call_id": "call_123",
        "content": get_weather("Beijing", "celsius")
    }
]

final_response = client.chat.completions.create(
    model="deepseek-ai/DeepSeek-V3.2-Exp",
    messages=messages,
    temperature=0.7
)

print(final_response.choices[0].message.content)
# Output: "The weather in Beijing is currently 22°C and sunny."
```

#### 4.2.3 Enabling PP, CP and TP with FP8 KV cache

We suggested `DP2` + `MTP` for local deployment of agentic workflow with DeepSeek V3.2 on Hopper platform:

```shell Command theme={null}
export SGLANG_DEEPEP_LL_COMBINE_SEND_NUM_SMS=32
export SGLANG_SET_CPU_AFFINITY=1

# Test workload ISL/OSL=1k/1k, raw tap : 4948.16 toks/sec, MAX ITL 5970
#   dp 2 : 5019.54  toks/sec, MAX ITL 7233
#   dp 4 : 4942.82  toks/sec, MAX ITL 35654
#   dp 2 + mtp : 6842.51 toks/sec, MAX ITL 3081
sglang_args=$(echo serve \
  --model-path $MAPPED_MODEL_PATH \
  --nccl-init $MASTER_ADDR:$MASTER_PORT --nnodes 2 --node-rank $RANK --tp 16 \
  --dp 2 --enable-dp-attention --page-size 64 \
  --trust-remote-code --host "0.0.0.0" --port 30000 \
  --log-requests \
  --context-length 65536 --max-running-requests 128 \
  --speculative-algorithm EAGLE \
  --speculative-num-steps 2 --speculative-eagle-topk 1 --speculative-num-draft-tokens 3 \
  --allow-auto-truncate --enable-metrics \
  --tool-call-parser deepseekv32 --reasoning-parser deepseek-v3 \
  --served-model-name DeepSeek-V3.2-Opt-dp2-mtp
)

sglang_args=($sglang_args)

sglang "${sglang_args[@]}" 2>&1 | tee $LOG_DIR/$RANK.log
```

**CP + PP + EP + DP**

`CP` is currently enabled with `PP=2` on Hopper platform and we can reduce TP=16 to TP=8 from standalone deployment:

```shell Command theme={null}
# verified on Hopper platform
sglang_args=$(echo serve \
  --model-path $MAPPED_MODEL_PATH \
  --nccl-init $MASTER_ADDR:$MASTER_PORT --nnodes 2 --node-rank $RANK --tp 8 --pp-size 2 --dp 1 --enable-dp-attention \
  --moe-a2a-backend deepep --ep-size 16  \
  --page-size 128 \
  --chunked-prefill-size 16384 \
  --attention-backend nsa \
  --nsa-prefill-backend flashmla_sparse \
  --nsa-decode-backend flashmla_sparse \
  --enable-nsa-prefill-context-parallel \
  --nsa-prefill-cp-mode round-robin-split \
  --cuda-graph-max-bs 128 \
  --max-running-requests 128 \
  --trust-remote-code --host "0.0.0.0" --port 30000 \
  --log-requests \
  --context-length 65536 \
  --allow-auto-truncate --enable-metrics \
  --tool-call-parser deepseekv32 --reasoning-parser deepseek-v3 \
  --served-model-name DeepSeek-V3.2-nsa-pp-cp-ep-dp
)

sglang_args=($sglang_args)

sglang "${sglang_args[@]}" 2>&1 | tee $LOG_DIR/$RANK.log
```

**fp8 KV + CP + PP**

With FP8 KV, we can have less memory footprint. This can be combined with various parallel schemes:

```shell Command theme={null}
# verified in Hopper platform
dp=1

dp_config=" \
  --dp 1 --enable-dp-attention \
"

cp_config=" \
  --enable-nsa-prefill-context-parallel \
"

if [ "$dp" -eq 1 ]; then

cp_config=" \
  $cp_config \
  --nsa-prefill-cp-mode round-robin-split \
"

else
cp_config=" \
  $cp_config \
  --nsa-prefill-cp-mode in-seq-split \
"
fi

# see discussion : https://github.com/sgl-project/sglang/pull/12065
sglang_args=$(echo serve \
  --model-path $MAPPED_MODEL_PATH \
  --nccl-init $MASTER_ADDR:$MASTER_PORT --nnodes 2 --node-rank $RANK --tp 8 --pp-size 2 --pp-async-batch-depth 1 \
  $dp_config \
  --trust-remote-code --host "0.0.0.0" --port 30000 \
  --log-requests \
  --context-length 65536 --max-running-requests 128 \
  $cp_config \
  --kv-cache-dtype fp8_e4m3 \
  --allow-auto-truncate --enable-metrics \
  --tool-call-parser deepseekv32 --reasoning-parser deepseek-v3 \
  --served-model-name DeepSeek-V3.2-Opt-fp8kv-pp2-cp4
)

sglang_args=($sglang_args)

sglang "${sglang_args[@]}" 2>&1 | tee $LOG_DIR/$RANK.log
```

## 5. Benchmark

### 5.1 Speed Benchmark on Blackwell

**Test Environment:**

* Hardware: NVIDIA B200 GPU (8x)
* Model: DeepSeek-V3.2-Exp
* Tensor Parallelism: 8
* sglang version: 0.5.6

We use SGLang's built-in benchmarking tool to conduct performance evaluation on the [ShareGPT\_Vicuna\_unfiltered](https://huggingface.co/datasets/anon8231489123/ShareGPT_Vicuna_unfiltered) dataset. This dataset contains real conversation data and can better reflect performance in actual use scenarios. To simulate real-world usage patterns, we configure each request with 1024 input tokens and 1024 output tokens, representing typical medium-length conversations with detailed responses.

#### 5.1.1 Latency-Sensitive Benchmark

* Model Deployment Command:

```shell Command theme={null}
sglang serve \
  --model-path deepseek-ai/DeepSeek-V3.2-Exp \
  --tp 8 \
  --speculative-algorithm EAGLE \
  --speculative-num-steps 3 \
  --speculative-eagle-topk 1 \
  --speculative-num-draft-tokens 4 \
  --host 0.0.0.0 \
  --port 30000
```

* Benchmark Command:

```shell Command theme={null}
python3 -m sglang.bench_serving \
  --backend sglang \
  --host 127.0.0.1 \
  --port 30000 \
  --model deepseek-ai/DeepSeek-V3.2-Exp \
  --random-input-len 1024 \
  --random-output-len 1024 \
  --num-prompts 10 \
  --max-concurrency 1
```

* **Test Results:**

```text Output theme={null}
============ Serving Benchmark Result ============
Backend:                                 sglang
Traffic request rate:                    inf
Max request concurrency:                 1
Successful requests:                     10
Benchmark duration (s):                  29.11
Total input tokens:                      1972
Total input text tokens:                 1972
Total input vision tokens:               0
Total generated tokens:                  2784
Total generated tokens (retokenized):    2777
Request throughput (req/s):              0.34
Input token throughput (tok/s):          67.73
Output token throughput (tok/s):         95.62
Peak output token throughput (tok/s):    157.00
Peak concurrent requests:                3
Total token throughput (tok/s):          163.36
Concurrency:                             1.00
Accept length:                           2.46
----------------End-to-End Latency----------------
Mean E2E Latency (ms):                   2909.74
Median E2E Latency (ms):                 3088.27
P90 E2E Latency (ms):                    4200.62
P99 E2E Latency (ms):                    5588.52
---------------Time to First Token----------------
Mean TTFT (ms):                          317.58
Median TTFT (ms):                        191.31
P99 TTFT (ms):                           740.79
-----Time per Output Token (excl. 1st token)------
Mean TPOT (ms):                          9.09
Median TPOT (ms):                        9.25
P99 TPOT (ms):                           11.73
---------------Inter-Token Latency----------------
Mean ITL (ms):                           9.35
Median ITL (ms):                         7.64
P95 ITL (ms):                            22.81
P99 ITL (ms):                            23.33
Max ITL (ms):                            31.45
==================================================
```

#### 5.1.2 Throughput-Sensitive Benchmark

* Model Deployment Command:

```shell Command theme={null}
sglang serve \
  --model-path deepseek-ai/DeepSeek-V3.2-Exp \
  --tp 8 \
  --ep 8 \
  --dp 8 \
  --enable-dp-attention \
  --host 0.0.0.0 \
  --port 30000
```

* Benchmark Command:

```shell Command theme={null}
python3 -m sglang.bench_serving \
  --backend sglang \
  --host 127.0.0.1 \
  --port 30000 \
  --model deepseek-ai/DeepSeek-V3.2-Exp \
  --random-input-len 1024 \
  --random-output-len 1024 \
  --num-prompts 1000 \
  --max-concurrency 100
```

* **Test Results:**

```text Output theme={null}
============ Serving Benchmark Result ============
Backend:                                 sglang
Traffic request rate:                    inf
Max request concurrency:                 100
Successful requests:                     1000
Benchmark duration (s):                  219.09
Total input tokens:                      301701
Total input text tokens:                 301701
Total input vision tokens:               0
Total generated tokens:                  188375
Total generated tokens (retokenized):    187443
Request throughput (req/s):              4.56
Input token throughput (tok/s):          1377.06
Output token throughput (tok/s):         859.80
Peak output token throughput (tok/s):    2465.00
Peak concurrent requests:                109
Total token throughput (tok/s):          2236.86
Concurrency:                             88.05
----------------End-to-End Latency----------------
Mean E2E Latency (ms):                   19291.23
Median E2E Latency (ms):                 11927.39
---------------Time to First Token----------------
Mean TTFT (ms):                          530.36
Median TTFT (ms):                        444.00
P99 TTFT (ms):                           1504.78
-----Time per Output Token (excl. 1st token)------
Mean TPOT (ms):                          106.16
Median TPOT (ms):                        106.69
P99 TPOT (ms):                           221.12
---------------Inter-Token Latency----------------
Mean ITL (ms):                           100.46
Median ITL (ms):                         41.73
P95 ITL (ms):                            225.67
P99 ITL (ms):                            392.37
Max ITL (ms):                            975.03
==================================================
```

### 5.2 Accuracy Benchmark

#### 5.2.1 GSM8K Benchmark

* **Benchmark Command:**

```shell Command theme={null}
python3 -m sglang.test.few_shot_gsm8k --num-questions 200 --port 30000
```

* **Test Results**:
  * DeepSeek-V3.2-Exp
    ```
    Accuracy: 0.980
    Invalid: 0.000
    Latency: 19.128 s
    Output throughput: 965.919 token/s
    ```

#### 5.2.2 MMLU Benchmark

* **Benchmark Command:**

```shell Command theme={null}
cd sglang
bash benchmark/mmlu/download_data.sh
python3 benchmark/mmlu/bench_sglang.py --nsub 10 --port 30000
```

* **Test Results**:
  * DeepSeek-V3.2-Exp
    ```
    subject: abstract_algebra, #q:100, acc: 0.780
    subject: anatomy, #q:135, acc: 0.874
    subject: astronomy, #q:152, acc: 0.961
    subject: business_ethics, #q:100, acc: 0.860
    subject: clinical_knowledge, #q:265, acc: 0.925
    subject: college_biology, #q:144, acc: 0.972
    subject: college_chemistry, #q:100, acc: 0.660
    subject: college_computer_science, #q:100, acc: 0.880
    subject: college_mathematics, #q:100, acc: 0.840
    subject: college_medicine, #q:173, acc: 0.879
    Total latency: 7.961
    Average accuracy: 0.879
    ```

### 5.3 Speed Benchmark on Hopper

**Test Environment:**

* Hardware: NVIDIA H800 GPU (16x)
* Model: DeepSeek-V3.2
* Tensor Parallelism: 16
* sglang version: 0.5.9

#### 5.3.1 Latency-Sensitive Benchmark

* Model Deployment Command:

```shell Command theme={null}
export SGLANG_DEEPEP_LL_COMBINE_SEND_NUM_SMS=32
export SGLANG_SET_CPU_AFFINITY=1

# Test workload ISL/OSL=1k/1k, raw tap : 4948.16 toks/sec, MAX ITL 5970
#   dp 2 : 5019.54  toks/sec, MAX ITL 7233
#   dp 4 : 4942.82  toks/sec, MAX ITL 35654
#   dp 2 + mtp : 6842.51 toks/sec, MAX ITL 3081
sglang_args=$(echo serve \
  --model-path $MAPPED_MODEL_PATH \
  --nccl-init $MASTER_ADDR:$MASTER_PORT --nnodes 2 --node-rank $RANK --tp 16 \
  --dp 2 --enable-dp-attention --page-size 64 \
  --trust-remote-code --host "0.0.0.0" --port 30000 \
  --log-requests \
  --context-length 65536 --max-running-requests 128 \
  --speculative-algorithm EAGLE \
  --speculative-num-steps 2 --speculative-eagle-topk 1 --speculative-num-draft-tokens 3 \
  --allow-auto-truncate --enable-metrics \
  --tool-call-parser deepseekv32 --reasoning-parser deepseek-v3 \
  --served-model-name DeepSeek-V3.2-Opt-dp2-mtp
)

sglang_args=($sglang_args)

sglang "${sglang_args[@]}" 2>&1 | tee $LOG_DIR/$RANK.log
```

* Benchmark Command:

```shell Command theme={null}
python3 -m sglang.bench_serving \
  --backend sglang \
  --host $MASTER_ADDR \
  --port 30000 \
  --model deepseek-ai/DeepSeek-V3.2 \
  --random-input-len 1024 \
  --random-output-len 1024 \
  --num-prompts 10 \
  --max-concurrency 1
```

* **Test Results:**

```text Output theme={null}
============ Serving Benchmark Result ============
Backend:                                 sglang
Traffic request rate:                    64.0
Max request concurrency:                 1
Successful requests:                     10
Benchmark duration (s):                  48.96
Total input tokens:                      6101
Total input text tokens:                 6101
Total generated tokens:                  4220
Total generated tokens (retokenized):    4217
Request throughput (req/s):              0.20
Input token throughput (tok/s):          124.62
Output token throughput (tok/s):         86.20
Peak output token throughput (tok/s):    113.00
Peak concurrent requests:                2
Total token throughput (tok/s):          210.81
Concurrency:                             1.00
Accept length:                           3.27
----------------End-to-End Latency----------------
Mean E2E Latency (ms):                   4893.12
Median E2E Latency (ms):                 3742.47
P90 E2E Latency (ms):                    8877.37
P99 E2E Latency (ms):                    10769.85
---------------Time to First Token----------------
Mean TTFT (ms):                          199.88
Median TTFT (ms):                        176.15
P99 TTFT (ms):                           272.49
-----Time per Output Token (excl. 1st token)------
Mean TPOT (ms):                          10.99
Median TPOT (ms):                        10.88
P99 TPOT (ms):                           13.93
---------------Inter-Token Latency----------------
Mean ITL (ms):                           11.15
Median ITL (ms):                         8.86
P95 ITL (ms):                            17.29
P99 ITL (ms):                            33.71
Max ITL (ms):                            36.84
==================================================
```

#### 5.3.2 Throughput-Sensitive Benchmark

We simply use the same deployment method and vary the throughput by maximizing concurrencies:

```shell Command theme={null}
python3 -m sglang.bench_serving \
  --backend sglang \
  --host $MASTER_ADDR \
  --port 30000 \
  --model deepseek-ai/DeepSeek-V3.2 \
  --random-input-len 1024 \
  --random-output-len 1024 \
  --num-prompts 2048 \
  --max-concurrency 1024 # see picture below why we use 1024 for concurrency, hence num prompts 2048
```

DeepSeek 3.2 can steadily support concurrency up to `1024` and when concurrency is greater than `128`, the TTFT increase sharply:

![DeepSeek V3.2 Concurrency ISL/OSL=1024/128](https://github.com/user-attachments/assets/d5c9c9fb-44f3-4793-a0fd-f8fa954546f5)

Performance record:

```text Output theme={null}
============ Serving Benchmark Result ============
Backend:                                 sglang
Traffic request rate:                    64.0
Max request concurrency:                 1024
Successful requests:                     2048
Benchmark duration (s):                  408.09
Total input tokens:                      1048992
Total input text tokens:                 1048992
Total generated tokens:                  1032734
Total generated tokens (retokenized):    1031817
Request throughput (req/s):              5.02
Input token throughput (tok/s):          2570.50
Output token throughput (tok/s):         2530.66
Peak output token throughput (tok/s):    5092.00
Peak concurrent requests:                1035
Total token throughput (tok/s):          5101.16
Concurrency:                             763.41
Accept length:                           3.26
----------------End-to-End Latency----------------
Mean E2E Latency (ms):                   152117.70
Median E2E Latency (ms):                 181704.84
P90 E2E Latency (ms):                    215924.77
P99 E2E Latency (ms):                    231679.59
---------------Time to First Token----------------
Mean TTFT (ms):                          127729.28
Median TTFT (ms):                        170098.94
P99 TTFT (ms):                           185705.73
-----Time per Output Token (excl. 1st token)------
Mean TPOT (ms):                          49.18
Median TPOT (ms):                        48.48
P99 TPOT (ms):                           77.24
---------------Inter-Token Latency----------------
Mean ITL (ms):                           48.46
Median ITL (ms):                         52.11
P95 ITL (ms):                            110.26
P99 ITL (ms):                            200.63
Max ITL (ms):                            2666.37
==================================================
```

By adding `--random-range-ratio 1`, we could get even higher statistical numbers:

```text Output theme={null}
============ Serving Benchmark Result ============
Backend:                                 sglang
Traffic request rate:                    64.0
Max request concurrency:                 1024
Successful requests:                     2048
Benchmark duration (s):                  612.87
Total input tokens:                      2097152
Total input text tokens:                 2097152
Total generated tokens:                  2097152
Total generated tokens (retokenized):    2096201
Request throughput (req/s):              3.34
Input token throughput (tok/s):          3421.84
Output token throughput (tok/s):         3421.84
Peak output token throughput (tok/s):    9077.00
Peak concurrent requests:                1039
Total token throughput (tok/s):          6843.68
Concurrency:                             772.66
Accept length:                           3.26
----------------End-to-End Latency----------------
Mean E2E Latency (ms):                   231222.27
Median E2E Latency (ms):                 289846.24
P90 E2E Latency (ms):                    314480.41
P99 E2E Latency (ms):                    320392.27
---------------Time to First Token----------------
Mean TTFT (ms):                          194081.02
Median TTFT (ms):                        252945.22
P99 TTFT (ms):                           279637.50
-----Time per Output Token (excl. 1st token)------
Mean TPOT (ms):                          36.31
Median TPOT (ms):                        36.73
P99 TPOT (ms):                           46.33
---------------Inter-Token Latency----------------
Mean ITL (ms):                           36.31
Median ITL (ms):                         23.18
P95 ITL (ms):                            96.79
P99 ITL (ms):                            135.81
Max ITL (ms):                            3121.00
==================================================
```