DeepSeek
Open weights3 channels

deepseek-v4-pro

DeepSeek V4 Pro is a large-scale Mixture-of-Experts model from DeepSeek with 1.6T total parameters and 49B activated parameters, supporting a 1M-token context window. It is designed for advanced reaso

Starting atSave 25%
Input
$1.2686/ 1M
List price $1.7143
Output
$2.5372/ 1M
List price $3.4286
Open dashboard

Available from 3 channels

Each channel is compared at its lowest input-price tier; output breaks ties. A dash means /api/pricing does not provide that rate. Selecting a channel updates the endpoints and call example below.

jd/deepseek-v4-pro
Lowest input
Input
$1.2686 USD per 1M tokens
Output
$2.5372 USD per 1M tokens
Cache read
$0.1057 USD per 1M tokens
Cache write
- USD per 1M tokens
baidu/deepseek-v4-pro
Input
$1.30 USD per 1M tokens
Output
$2.60 USD per 1M tokens
Cache read
$0.1077 USD per 1M tokens
Cache write
- USD per 1M tokens
deepseek/deepseek-v4-pro
Input
$1.32 USD per 1M tokens
Output
$3.96 USD per 1M tokens
Cache read
$0.044 USD per 1M tokens
Cache write
- USD per 1M tokens

Endpoints

POST/v1/chat/completionsChat Completions

Call it

Using jd/deepseek-v4-pro

1from openai import OpenAI23client = OpenAI(4    base_url="https://www.realrelay.ai/v1",5    api_key="sk-***",6)78response = client.chat.completions.create(9    model="jd/deepseek-v4-pro",10    messages=[{"role": "user", "content": "Hello"}],11)1213print(response.choices[0].message.content)
1import OpenAI from "openai";23const client = new OpenAI({4  baseURL: "https://www.realrelay.ai/v1",5  apiKey: process.env.REALRELAY_API_KEY,6});78const response = await client.chat.completions.create({9  model: "jd/deepseek-v4-pro",10  messages: [{ role: "user", content: "Hello" }],11});1213console.log(response.choices[0].message.content);
1curl https://www.realrelay.ai/v1/chat/completions \2  -H "Authorization: Bearer $REALRELAY_API_KEY" \3  -H "Content-Type: application/json" \4  -d '{5    "model": "jd/deepseek-v4-pro",6    "messages": [{ "role": "user", "content": "Hello" }]7  }'

Prices and availability as published on Oct 4, 2026.

Capabilities

MCPPrompt cachingReasoningTool useWeb search
Input
Text
Output
Text