DeepSeek-V4-Pro

deepseek

DeepSeek V4 Pro is an MoE model with 1.6T total parameters (49B active) and a 1M-token context window. It's built for advanced reasoning, coding, and long-running agent tasks, and performs well on knowledge, math, and software engineering benchmarks.

Context
1.0M
Max output
384K
Input price
$1.3/1M tokens
Output price
$2.6/1M tokens

Capabilities

Reasoning & Interaction
Reasoning
Tool Calling
Output Control
Structured Output
JSON Mode
Infrastructure
Streaming
Fine-tuning

Details

Model IDdeepseek-ai/DeepSeek-V4-Pro
Provider DeepInfra
Creator DeepSeek
Familydeepseek-v4
Licensedeepseek
Statusactive
Typechat
Input modalitiestext
Output modalitiestext
Open weightYes
Parameters—
Architecture—
Context window1.0M tokens
Max input—
Max output384K tokens
Reasoning tokensYes
Knowledge cutoff—
Release date
Sourceofficial

Tools

Function Callingfunction_calling
Call external functions and APIs

Endpoints

Chat CompletionsPOST
Generate chat responses with messageshttps://api.deepinfra.com/v1/openai/v1/chat/completions

Pricing

Input
$1.3
Output
$2.6
Cache write
—
Cache read
$0.1
Batch in
—
Batch out
—

Family Comparison: deepseek-v4

ModelContextMax outPricing
DeepSeek-V4-Flash1.0M384K$0.09$0.18
DeepSeek-V4-Pro1.0M384K$1.3$2.6

API

GET/v1/models/deepinfra/deepseek-ai/DeepSeek-V4-Pro