> For clean Markdown of any page, append .md to the page URL. > For a complete documentation index, see https://docs.empiriolabs.ai/llms.txt. > For AI client integration (Claude Code, Cursor, etc.), connect to the MCP server at https://docs.empiriolabs.ai/_mcp/server. # DeepSeek V4 Pro > Flagship MoE LLM with 1.6T total / 49B active parameters and native 1M context for advanced math, logical inference, and specialized coding. ![DeepSeek V4 Pro](https://media.empiriolabs.ai/model-logos/deepseek.png) [DeepSeek](/providers/deepseek) · Text Generation `POST /v1/chat/completions` Flagship MoE LLM with 1.6T total / 49B active parameters and native 1M context for advanced math, logical inference, and specialized coding. ## At a glance | Field | Value | | ------------------- | ----------------------------------------------------------------------------------------------------------------------------- | | Model id | `deepseek-v4-pro` | | Model release date | 2026-04-24 | | Input modalities | Text | | Output modalities | Text | | Context window | 1M | | Weight precision | - | | Max output tokens | 393,216 | | Region | Germany | | Features | reasoning, function_calling | | Native inference | No | | New | No | | Structured output | JSON Schema | | Supported endpoints | `POST /v1/chat/completions`, `POST /v1/responses`, `POST /v1/messages`, `POST /v1beta/models/deepseek-v4-pro:generateContent` | | Alternate model ids | `deepseek/deepseek-v4-pro` | ## Pricing | Charge | Spec | Rate | | ------------------- | ----------------------- | ----------------- | | Input | per 1M prompt tokens | ~~\$1.74~~ \$1.65 | | Output | per 1M generated tokens | ~~\$3.48~~ \$3.30 | | Web Search (Linkup) | per call when invoked | \$0.013 | ## Example request ```bash curl https://api.empiriolabs.ai/v1/chat/completions \ -H 'Authorization: Bearer $EMPIRIOLABS_API_KEY' \ -H 'Content-Type: application/json' \ -d '{"model": "deepseek-v4-pro", "messages": [{"role":"user","content":"Hello"}]}' ``` ## Parameters | Parameter | Type | Required | Default | Description | | -------------------- | ------- | -------- | ---------- | ----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- | | `temperature` | number | no | `0.7` | Sampling temperature. 0 = deterministic, 2 = maximum randomness. · Range: 0 – 2 | | `top_p` | number | no | `0.9` | Nucleus sampling probability mass. Lower = more focused. · Range: 0 – 1 | | `max_tokens` | number | no | - | Maximum output tokens. When omitted, the model's own default applies. · Range: 1 – 393216 | | `stop` | string | no | - | Up to 4 strings where the model will stop generating further tokens. | | `enable_thinking` | boolean | no | true | Enable step-by-step reasoning before answering. | | `thinking_budget` | number | no | `32768` | Maximum tokens reserved for reasoning when thinking is enabled. · Range: 1 – 393216 | | `reasoning_effort` | enum | no | `"medium"` | Reasoning effort level. none disables thinking. low, medium, high, and max set bounded thinking budgets sized to the selected model. · Allowed: `none`, `low`, `medium`, `high`, `max` | | `response_format` | enum | no | - | Constrain the output to JSON. Use JSON mode for any valid JSON object, or JSON schema to force output that matches a schema you provide. | | `web_search_linkup` | boolean | no | false | Optional web search powered by Linkup. When enabled, recent web sources are retrieved using your latest user message as the query and provided to the model as additional context. Adds \$0.013 per call when invoked on top of the model's normal token cost. Disabled by default. | | `disable_formatting` | boolean | no | false | Return the answer without the Sources list that web search adds. | ## Variants ### `:variant1` | Field | Value | | ------------------- | -------------------------------------------------------------------------------------------------------------------------------------- | | Model id | `deepseek-v4-pro:variant1` | | Model release date | 2026-04-24 | | Region | Singapore | | Context window | 1M | | Weight precision | - | | Max output tokens | 393,216 | | Features | reasoning, web_search, function_calling | | Native inference | No | | Structured output | JSON Schema | | Supported endpoints | `POST /v1/chat/completions`, `POST /v1/responses`, `POST /v1/messages`, `POST /v1beta/models/deepseek-v4-pro:variant1:generateContent` | **Pricing** | Charge | Spec | Rate | | ---------- | ------------------------ | ------ | | Input | per 1M prompt tokens | \$2.40 | | Output | per 1M generated tokens | \$4.80 | | Web search | per request when enabled | \$0.02 | **Parameters** | Parameter | Type | Required | Default | Description | | ------------------ | ------- | -------- | ---------- | -------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- | | `temperature` | number | no | `0.7` | Sampling temperature. 0 = deterministic, 2 = maximum randomness. · Range: 0 – 2 | | `top_p` | number | no | `0.9` | Nucleus sampling probability mass. Lower = more focused. · Range: 0 – 1 | | `max_tokens` | number | no | - | Maximum output tokens. When omitted, the model's own default applies. · Range: 1 – 393216 | | `stop` | string | no | - | Up to 4 strings where the model will stop generating further tokens. | | `enable_thinking` | boolean | no | true | Enable step-by-step reasoning before answering. | | `thinking_budget` | number | no | `32768` | Maximum tokens reserved for reasoning when thinking is enabled. · Range: 1 – 393216 | | `reasoning_effort` | enum | no | `"medium"` | Reasoning effort level. none disables thinking. low, medium, high, and max set bounded thinking budgets sized to the selected model. · Allowed: `none`, `low`, `medium`, `high`, `max` | | `tool_web_search` | boolean | no | false | Enable live web search. Adds \$0.02 to the request cost when enabled. | | `response_format` | enum | no | - | Constrain the output to JSON. Use JSON mode for any valid JSON object, or JSON schema to force output that matches a schema you provide. | ### `:variant2` | Field | Value | | ------------------- | -------------------------------------------------------------------------------------------------------------------------------------- | | Model id | `deepseek-v4-pro:variant2` | | Model release date | 2026-04-24 | | Region | China | | Context window | 1M | | Weight precision | - | | Max output tokens | 393,216 | | Features | reasoning, function_calling, web_search, cache | | Native inference | No | | Structured output | JSON Schema | | Supported endpoints | `POST /v1/chat/completions`, `POST /v1/responses`, `POST /v1/messages`, `POST /v1beta/models/deepseek-v4-pro:variant2:generateContent` | **Pricing** | Charge | Spec | Rate | | ------------------- | -------------------------- | ------------------ | | Input | per 1M prompt tokens | ~~\$1.74~~ \$1.65 | | Output | per 1M generated tokens | ~~\$3.48~~ \$3.301 | | Implicit cache read | per 1M cached input tokens | \$0.138 | | Web search | per request when enabled | \$0.01 | **Parameters** | Parameter | Type | Required | Default | Description | | ------------------ | ------- | -------- | ---------- | -------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- | | `temperature` | number | no | `0.7` | Sampling temperature. 0 = deterministic, 2 = maximum randomness. · Range: 0 – 2 | | `top_p` | number | no | `0.9` | Nucleus sampling probability mass. Lower = more focused. · Range: 0 – 1 | | `max_tokens` | number | no | - | Maximum output tokens. When omitted, the model's own default applies. · Range: 1 – 393216 | | `stop` | string | no | - | Up to 4 strings where the model will stop generating further tokens. | | `enable_thinking` | boolean | no | true | Enable step-by-step reasoning before answering. With response_format, the model answers without reasoning unless the request sets enable_thinking. | | `thinking_budget` | number | no | `32768` | Maximum tokens reserved for reasoning when thinking is enabled. · Range: 1 – 393216 | | `reasoning_effort` | enum | no | `"medium"` | Reasoning effort level. none disables thinking. low, medium, high, and max set bounded thinking budgets sized to the selected model. · Allowed: `none`, `low`, `medium`, `high`, `max` | | `tool_web_search` | boolean | no | false | Enable live web search. Adds \$0.01 to the request cost when enabled. | | `response_format` | enum | no | - | Constrain the output to JSON. Use JSON mode for any valid JSON object, or JSON schema to force output that matches a schema you provide. | ### `:variant3` | Field | Value | | ------------------- | -------------------------------------------------------------------------------------------------------------------------------------- | | Model id | `deepseek-v4-pro:variant3` | | Model release date | 2026-04-24 | | Region | United States | | Context window | 1M | | Weight precision | - | | Max output tokens | 393,216 | | Features | reasoning, function_calling, structured_output | | Native inference | No | | Structured output | JSON Schema | | Supported endpoints | `POST /v1/chat/completions`, `POST /v1/responses`, `POST /v1/messages`, `POST /v1beta/models/deepseek-v4-pro:variant3:generateContent` | **Pricing** | Charge | Spec | Rate | | ------------------- | ----------------------- | ------- | | Input | per 1M prompt tokens | \$2.40 | | Output | per 1M generated tokens | \$4.80 | | Web Search (Linkup) | per call when invoked | \$0.013 | **Parameters** | Parameter | Type | Required | Default | Description | | -------------------- | ------- | -------- | ---------- | ----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- | | `temperature` | number | no | `0.7` | Sampling temperature. 0 = deterministic, 2 = maximum randomness. · Range: 0 – 2 | | `top_p` | number | no | `0.9` | Nucleus sampling probability mass. Lower = more focused. · Range: 0 – 1 | | `max_tokens` | number | no | - | Maximum output tokens. When omitted, the model's own default applies. · Range: 1 – 393216 | | `stop` | string | no | - | Up to 4 strings where the model will stop generating further tokens. | | `enable_thinking` | boolean | no | true | Enable step-by-step reasoning before answering. | | `reasoning_effort` | enum | no | `"medium"` | Reasoning effort level. none disables thinking. low, medium, high, and max set bounded thinking budgets sized to the selected model. · Allowed: `none`, `low`, `medium`, `high`, `max` | | `thinking_budget` | number | no | `32768` | Maximum tokens reserved for reasoning when thinking is enabled. · Range: 1 – 393216 | | `response_format` | enum | no | - | Constrain the output to JSON. Use JSON mode for any valid JSON object, or JSON schema to force output that matches a schema you provide. | | `web_search_linkup` | boolean | no | false | Optional web search powered by Linkup. When enabled, recent web sources are retrieved using your latest user message as the query and provided to the model as additional context. Adds \$0.013 per call when invoked on top of the model's normal token cost. Disabled by default. | | `disable_formatting` | boolean | no | false | Return the answer without the Sources list that web search adds. | --- *Machine-readable schema:* `GET https://api.empiriolabs.ai/v1/models/deepseek-v4-pro`.