deepseek-v3

Updated at:

A self-developed Mixture-of-Experts (MoE) model with 671B parameters (activating 37B), pre-trained on 14.8T tokens. Demonstrates excellent capabilities in long-text processing, coding, mathematics, encyclopedic knowledge, and Chinese language tasks.

Inference Service Provider

The inference service provider for deepseek-v3 is Alibaba Cloud Model Studio.

Model Capabilities

CapabilitySupportCapabilitySupport

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

ParameterValueParameterValue

Max Input Length

57344

Max Output Length

8192

Context Window

65536

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 ItemPrice (CNY)Unit

Input

2

Per 1M tokens

Output

8

Per 1M tokens

Input(Implicit Cache)

0.4

Per 1M tokens

Input(Batch File)

1

Per 1M tokens

Output(Batch File)

4

Per 1M tokens

Rate Limits

China (Beijing)

ParameterValue

RPM (Requests Per Minute)

15000

TPM (Tokens Per Minute)

1,200,000