MODEL IDdeepseek-v4-flash

live

DeepSeek-V4-Flash

by DeepSeekApril 24, 2026

DeepSeek-V4-Flash is DeepSeek's fast, efficient, and cost-focused frontier language model for coding, reasoning, and agent workflows. It supports both thinking and non-thinking modes, a 1M token context window, up to 384K output tokens, tool calls, JSON output, and efficient long-context operation for software, research, and structured professional tasks.

API Options

Platform-level options for task execution and delivery.

taskType stringrequiredvalue: textInference: Identifier for the type of task being performed

taskUUID stringrequiredUUID v4: UUID v4 identifier for tracking tasks and matching async responses. Must be unique per task.

outputFormat stringdefault: TEXT

Specifies the file format of the generated output. The available values depend on the task type and the specific model's capabilities.

Allowed values2 values

webhookURL stringURI

Specifies a webhook URL where JSON responses will be sent via HTTP POST when generation tasks complete. For batch requests with multiple results, each completed item triggers a separate webhook call as it becomes available.

Learn more1 resource

Webhooks
PLATFORM

deliveryMethod stringdefault: sync

Determines how the API delivers task results.

Allowed values3 values

: Returns complete results directly in the API response.
: Returns an immediate acknowledgment with the task UUID. Poll for results using getResponse.
: Streams results token-by-token as they are generated.

Learn more1 resource

Task Polling
PLATFORM

includeCost booleandefault: false: Include task cost in the response.

includeUsage booleandefault: false: Include token usage statistics in the response.

Core Parameters

Primary parameters that define the task output.

numberResults integermin: 1max: 4default: 1: Number of results to generate. Each result uses a different seed, producing variations of the same parameters.

model stringrequiredvalue: deepseek-v4-flash: Identifier of the model to use for generation.

seed integermin: 0max: 9223372036854776000: Random seed for reproducible generation. When not provided, a random seed is generated in the unsigned 32-bit range.

jsonSchema object | string: JSON Schema for structured output. Only honoured when outputFormat is JSON. Accepts the OpenAI envelope ({name, schema, strict}) or a bare JSON Schema; bare schemas are auto-wrapped with name='response' and strict=true.

messages array of objectsrequiredmin items: 1

Array of chat messages forming the conversation context.

Properties2 properties

messages » role role stringrequired: The role of the message author.
Allowed values2 values

messages » content content stringrequiredmin: 1: The text content of the message.

tools array of objectsmin items: 1

Tool definitions available for the model to call during generation.

Properties4 properties

tools » type type stringrequired: The kind of tool to make available to the model. User-defined functions require name and schema, while built-in tools (search, codeInterpreter) are executed server-side by the provider.
Allowed values1 value

tools » name name stringmax: 64: Unique function name. Required for function tools.

tools » description description string: Explanation of what the function does, used by the model to decide when to call it.

tools » schema schema object: JSON Schema object describing the function's input parameters.

Settings

Technical parameters to fine-tune the inference process. These must be nested inside the settings object.

settings » systemPrompt systemPrompt stringmin: 1max: 1000000: System-level instruction that guides the model's behavior and output style across the entire generation.

settings » temperature temperature floatmin: 0max: 2step: 0.01default: 1: Controls randomness in generation. Lower values produce more deterministic outputs, higher values increase variation and creativity.

settings » topP topP floatmin: 0max: 1step: 0.01default: 1: Nucleus sampling parameter that controls diversity by limiting the probability mass. Lower values make outputs more focused, higher values increase diversity.

settings » frequencyPenalty frequencyPenalty floatmin: 0max: 2step: 0.01default: 0: Penalizes tokens based on their frequency in the output so far. A value of 0.0 disables the penalty.

settings » maxTokens maxTokens integermin: 1max: 1048576default: 32768: Maximum number of tokens to generate in the response.

settings » presencePenalty presencePenalty floatmin: 0max: 2step: 0.01default: 0: Encourages the model to introduce new topics. A value of 0.0 disables the penalty.

settings » stopSequences stopSequences array of stringsmin: 1max: 50max items: 5: Array of sequences that will cause the model to stop generating further tokens when encountered.

settings » thinkingLevel thinkingLevel stringdefault: off: Controls the depth of internal reasoning the model performs before generating a response.
Allowed values3 values

toolChoice object

Controls how the model selects which tool to call. This only takes effect when tools are defined.

Examples3 examples

Let the model decide (default):

"toolChoice": {
  "type": "auto"
}

Force a specific tool call:

"toolChoice": {
  "type": "tool",
  "name": "get_weather"
}

Require any tool call:

"toolChoice": {
  "type": "any"
}

Properties2 properties

toolChoice » type type stringrequired

Strategy the model uses to decide when and which tools to call.

Allowed values4 values

: The model decides whether to call a tool based on the conversation context. This is the recommended default.
: The model must call at least one tool but chooses which one. Useful when you always need structured output.
: The model must call the specific tool identified by name. Use this to force a particular function call.
: The model will not call any tool, even if tools are defined. Useful for forcing a text-only response.

toolChoice » name name string: Name of the specific tool the model must call. Required when type is tool.