deepseek-v4-flash

更新时间:
复制 MD 格式

A highly efficient, lightweight MoE model with 284 billion parameters in total and 13 billion activated parameters, natively supporting context windows of up to one million tokens. It offers fast inference speed, low latency, and cost-effective invocation, delivering well-balanced overall performance. Designed for high-concurrency, lightweight workloads, it is ideally suited for common, essential use cases such as everyday dialogue, content creation, basic RAG applications, and batch text processing. This model corresponds to DeepSeek's preview release (deepseek-v4-flash-preview). For the official stable version, use deepseek-v4-flash-0731.

Inference Service Provider

The inference service provider for deepseek-v4-flash is Alibaba Cloud Model Studio.

Model Capabilities

China (Beijing)

Capability Support Capability Support

Input Modality

Text

Output Modality

Text

Model Experience

Supported

Function Calling

Supported

Structured Outputs

Unsupported

Web Search

Supported

Prefix Completion

Unsupported

Context Caching

Supported

Batch Inference

Unsupported

Fine-tuning

Unsupported

Singapore

Scope: International

Capability Support Capability Support

Input Modality

Text

Output Modality

Text

Model Experience

Supported

Function Calling

Supported

Structured Outputs

Unsupported

Web Search

Supported

Prefix Completion

Unsupported

Context Caching

Supported

Batch Inference

Unsupported

Fine-tuning

Unsupported

Germany (Frankfurt)

Scope: Global

Capability Support Capability Support

Input Modality

Text

Output Modality

Text

Model Experience

Supported

Function Calling

Supported

Structured Outputs

Unsupported

Web Search

Unsupported

Prefix Completion

Unsupported

Context Caching

Supported

Batch Inference

Unsupported

Fine-tuning

Unsupported

US (Virginia)

Scope: Global

Capability Support Capability Support

Input Modality

Text

Output Modality

Text

Model Experience

Supported

Function Calling

Supported

Structured Outputs

Unsupported

Web Search

Unsupported

Prefix Completion

Unsupported

Context Caching

Supported

Batch Inference

Unsupported

Fine-tuning

Unsupported

Japan (Tokyo)

Scope: Global

Capability Support Capability Support

Input Modality

Text

Output Modality

Text

Model Experience

Supported

Function Calling

Supported

Structured Outputs

Unsupported

Web Search

Supported

Prefix Completion

Unsupported

Context Caching

Supported

Batch Inference

Unsupported

Fine-tuning

Unsupported

Context Limits

Parameter Value Parameter Value

Max Input Length

1000000

Max Output Length

393216

Context Window

1000000

Pricing

This page only shows the original pricing for model API calls, excluding any limited-time promotions. Visit Model Studio Console for promotional offers.

China (Beijing)

Billing Item Price (CNY) Unit

Input

1

Per 1M tokens

Output

2

Per 1M tokens

Input(Implicit Cache)

0.2

Per 1M tokens

Singapore

Scope: International

Billing Item Price (CNY) Unit

Input

1.499

Per 1M tokens

Output

2.998

Per 1M tokens

Input(Implicit Cache)

0.3

Per 1M tokens

Germany (Frankfurt)

Scope: Global

Billing Item Price (CNY) Unit

Input

1

Per 1M tokens

Output

2

Per 1M tokens

Input(Implicit Cache)

0.2

Per 1M tokens

US (Virginia)

Scope: Global

Billing Item Price (CNY) Unit

Input

1

Per 1M tokens

Output

2

Per 1M tokens

Input(Implicit Cache)

0.2

Per 1M tokens

Japan (Tokyo)

Scope: Global

Billing Item Price (CNY) Unit

Input

1

Per 1M tokens

Output

2

Per 1M tokens

Input(Implicit Cache)

0.2

Per 1M tokens

Rate Limits

China (Beijing)

Parameter Value

RPM (Requests Per Minute)

15000

TPM (Tokens Per Minute)

1,200,000

Singapore

Scope: International

Parameter Value

RPM (Requests Per Minute)

10000

TPM (Tokens Per Minute)

1,200,000

Germany (Frankfurt)

Scope: Global

Parameter Value

RPM (Requests Per Minute)

15000

TPM (Tokens Per Minute)

1,200,000

US (Virginia)

Scope: Global

Parameter Value

RPM (Requests Per Minute)

10000

TPM (Tokens Per Minute)

1,200,000

Japan (Tokyo)

Scope: Global

Parameter Value

RPM (Requests Per Minute)

10000

TPM (Tokens Per Minute)

1,200,000

Snapshot Versions

deepseek-v4-flash-0731

A highly efficient, lightweight MoE model with 284 billion parameters in total and 13 billion activated parameters, natively supporting context windows of up to one million tokens. It offers fast inference speed, low latency, and cost-effective invocation, delivering well-balanced overall performance. Designed for high-concurrency, lightweight workloads, it is ideally suited for common, essential use cases such as everyday dialogue, content creation, basic RAG applications, and batch text processing. This version corresponds to the official stable release of DeepSeek-V4-Flash (July 31, 2026) and is the recommended version for production use.

Model Capabilities

China (Beijing)

Capability Support Capability Support

Input Modality

Text

Output Modality

Text

Model Experience

Supported

Function Calling

Supported

Structured Outputs

Unsupported

Web Search

Supported

Partial Mode

Unsupported

Context Cache

Supported

Batch Inference

Unsupported

Fine-tuning

Unsupported

Singapore

Scope: International

Capability Support Capability Support

Input Modality

Text

Output Modality

Text

Model Experience

Supported

Function Calling

Supported

Structured Outputs

Unsupported

Web Search

Supported

Partial Mode

Unsupported

Context Cache

Supported

Batch Inference

Unsupported

Fine-tuning

Unsupported

Germany (Frankfurt)

Scope: Global

Capability Support Capability Support

Input Modality

Text

Output Modality

Text

Model Experience

Supported

Function Calling

Supported

Structured Outputs

Unsupported

Web Search

Supported

Partial Mode

Unsupported

Context Cache

Supported

Batch Inference

Unsupported

Fine-tuning

Unsupported

US (Virginia)

Scope: Global

Capability Support Capability Support

Input Modality

Text

Output Modality

Text

Model Experience

Supported

Function Calling

Supported

Structured Outputs

Unsupported

Web Search

Supported

Partial Mode

Unsupported

Context Cache

Supported

Batch Inference

Unsupported

Fine-tuning

Unsupported

Japan (Tokyo)

Scope: Global

Capability Support Capability Support

Input Modality

Text

Output Modality

Text

Model Experience

Supported

Function Calling

Supported

Structured Outputs

Unsupported

Web Search

Supported

Partial Mode

Unsupported

Context Cache

Supported

Batch Inference

Unsupported

Fine-tuning

Unsupported

Context Limits

Parameter Value Parameter Value

Max Input Length

1000000

Max Output Length

393216

Context Window

1000000

Pricing

This page only shows the original pricing for model API calls, excluding any limited-time promotions. Visit Model Studio Console for promotional offers.

China (Beijing)

Billing Item Price (CNY) Unit

Input

1

Per 1M tokens

Output

2

Per 1M tokens

Input(Implicit Cache)

0.2

Per 1M tokens

Singapore

Scope: International

Billing Item Price (CNY) Unit

Input

1.5

Per 1M tokens

Output

3

Per 1M tokens

Input(Implicit Cache)

0.3

Per 1M tokens

Germany (Frankfurt)

Scope: Global

Billing Item Price (CNY) Unit

Input

1

Per 1M tokens

Output

2

Per 1M tokens

Input(Implicit Cache)

0.2

Per 1M tokens

US (Virginia)

Scope: Global

Billing Item Price (CNY) Unit

Input

1

Per 1M tokens

Output

2

Per 1M tokens

Input(Implicit Cache)

0.2

Per 1M tokens

Japan (Tokyo)

Scope: Global

Billing Item Price (CNY) Unit

Input

1

Per 1M tokens

Output

2

Per 1M tokens

Input(Implicit Cache)

0.2

Per 1M tokens

Rate Limits

China (Beijing)

Parameter Value

RPM (Requests Per Minute)

15,000

TPM (Tokens Per Minute)

1,200,000

Singapore

Scope: International

Parameter Value

RPM (Requests Per Minute)

15,000

TPM (Tokens Per Minute)

1,200,000

Germany (Frankfurt)

Scope: Global

Parameter Value

RPM (Requests Per Minute)

15,000

TPM (Tokens Per Minute)

1,200,000

US (Virginia)

Scope: Global

Parameter Value

RPM (Requests Per Minute)

15,000

TPM (Tokens Per Minute)

1,200,000

Japan (Tokyo)

Scope: Global

Parameter Value

RPM (Requests Per Minute)

15,000

TPM (Tokens Per Minute)

1,200,000