DeepSeek 模型與 API

瀏覽 TokenHub 目前可用的 DeepSeek 模型，比較價格、上下文視窗與支援能力，並透過相容 OpenAI 的 API 工作流程完成串接。

查看 DeepSeek 模型取得 API Key

DeepSeek 模型與 API 定價

比較目前 DeepSeek API 價格、輸入與輸出 token 成本、上下文視窗和支援端點。開啟模型頁可查看完整規格與請求範例。

API 價格模型參數

獨立比較各模型的輸入、輸出與快取讀取價格，避免與能力參數混在一起。

模型名稱輸入/ 百萬 Token輸出/ 百萬 Token快取讀取/ 百萬 Token

DeepSeek

DeepSeek R1deepseek-r1

$0.5714$2.2857$0.2286

DeepSeek

DeepSeek V3deepseek-v3

$0.2857$1.1429$0.1143

DeepSeek

DeepSeek V3.1 Terminusdeepseek-v3.1-terminus

$0.5714$1.7143$0.0857

DeepSeek

DeepSeek V3.2 Thinkdeepseek-v3.2-think

$0.2857$0.4286$0.0286

DeepSeek

DeepSeek V4 Flashdeepseek-v4-flash

$0.4286$1.2857$0.0143

DeepSeek

DeepSeek V4 Flash Vision Expdeepseek-v4-flash-vision-exp

$0.4286$1.2857$0.0143

DeepSeek

DeepSeek V4 Prodeepseek-v4-pro

$1.2857$3.8571$0.0429

比較上下文視窗、最大輸出、推理、工具呼叫、端點和發布日期。

模型名稱模態上下文窗口最大輸出推理工具呼叫端點發布日期

DeepSeek

DeepSeek R1deepseek-r1

128K

32.8K

✓

OpenAI / Anthropic

2025年1月20日

DeepSeek

DeepSeek V3deepseek-v3

128K

16K

—

OpenAI / Anthropic

2024年12月26日

DeepSeek

DeepSeek V3.1 Terminusdeepseek-v3.1-terminus

128K

64K

—

OpenAI / Anthropic

2025年9月22日

DeepSeek

DeepSeek V3.2 Thinkdeepseek-v3.2-think

128K

64K

—

OpenAI / Anthropic

2025年12月1日

DeepSeek

DeepSeek V4 Flashdeepseek-v4-flash

384K

✓

OpenAI / Responses

2026年4月24日

DeepSeek

DeepSeek V4 Flash Vision Expdeepseek-v4-flash-vision-exp

384K

✓

Responses / Anthropic / OpenAI

2026年8月21日

DeepSeek

DeepSeek V4 Prodeepseek-v4-pro

384K

✓

OpenAI / Responses

2026年4月24日

目錄資料說明

價格、模型可用性、上下文視窗和端點支援來自 TokenHub 目前模型目錄。規格與可用性可能調整，生產環境上線前請在模型詳情頁再次確認。

最後更新: 2026-08-25

開始呼叫 DeepSeek API

透過 DeepSeek OpenAI API 相容方式呼叫 Chat Completions 或 Responses，也可以使用 DeepSeek Claude Messages API。頁面提供 OpenAI SDK 呼叫 DeepSeek 的 Python 和 Node.js 範例；是否支援串流輸出取決於所選端點。

01建立 TokenHub API Key
02選擇目前可用的 DeepSeek 模型
03選擇 API 協定
04使用發布的端點和模型 ID
05傳送第一個請求

取得 API Key 閱讀 API 文件

選擇 API 協定

OpenAI ChatPOST /v1/chat/completionsOpenAI ResponsesPOST /v1/responsesClaude MessagesPOST /v1/messages

OpenAI SDK · TypeScriptOpenAI SDK · PythoncURLTypeScript fetchPython requests

import OpenAI from "openai"

const client = new OpenAI({
  apiKey: process.env.TOKENHUB_API_KEY,
  baseURL: "https://us-api.tokenhub.com/v1",
})

const result = await client.chat.completions.create({
  "model": "deepseek-v3.2-think",
  "messages": [
    {
      "role": "user",
      "content": "Explain why low latency matters for an AI product in one sentence."
    }
  ]
})
console.log(result.choices[0]?.message?.content)

import json
import os
from openai import OpenAI

client = OpenAI(
    api_key=os.environ.get("TOKENHUB_API_KEY"),
    base_url="https://us-api.tokenhub.com/v1",
)

request = json.loads("{\"model\":\"deepseek-v3.2-think\",\"messages\":[{\"role\":\"user\",\"content\":\"Explain why low latency matters for an AI product in one sentence.\"}]}")
result = client.chat.completions.create(**request)
print(result.choices[0].message.content)

curl 'https://us-api.tokenhub.com/v1/chat/completions' \
  -X 'POST' \
  -H "Authorization: Bearer $TOKENHUB_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
  "model": "deepseek-v3.2-think",
  "messages": [
    {
      "role": "user",
      "content": "Explain why low latency matters for an AI product in one sentence."
    }
  ]
}'

const response = await fetch("https://us-api.tokenhub.com/v1/chat/completions", {
  method: "POST",
  headers: {
    Authorization: `Bearer ${process.env.TOKENHUB_API_KEY}`,
    "Content-Type": "application/json",
  },
  body: JSON.stringify({
    "model": "deepseek-v3.2-think",
    "messages": [
      {
        "role": "user",
        "content": "Explain why low latency matters for an AI product in one sentence."
      }
    ]
  }),
})
const data = await response.json()
console.log(data)

import os
import requests

response = requests.request(method="POST", url="https://us-api.tokenhub.com/v1/chat/completions",
    headers={
        "Authorization": f"Bearer {os.environ['TOKENHUB_API_KEY']}",
        "Content-Type": "application/json",
    },
    json=__import__("json").loads("{\"model\":\"deepseek-v3.2-think\",\"messages\":[{\"role\":\"user\",\"content\":\"Explain why low latency matters for an AI product in one sentence.\"}]}"),
)
response.raise_for_status()
print(response.json())

OpenAI SDK · TypeScriptOpenAI SDK · PythoncURLTypeScript fetchPython requests

import OpenAI from "openai"

const client = new OpenAI({
  apiKey: process.env.TOKENHUB_API_KEY,
  baseURL: "https://us-api.tokenhub.com/v1",
})

const result = await client.responses.create({
  "model": "deepseek-v3.2-think",
  "input": "Explain why low latency matters for an AI product in one sentence."
})
console.log(result.output_text)

import json
import os
from openai import OpenAI

client = OpenAI(
    api_key=os.environ.get("TOKENHUB_API_KEY"),
    base_url="https://us-api.tokenhub.com/v1",
)

request = json.loads("{\"model\":\"deepseek-v3.2-think\",\"input\":\"Explain why low latency matters for an AI product in one sentence.\"}")
result = client.responses.create(**request)
print(result.output_text)

curl 'https://us-api.tokenhub.com/v1/responses' \
  -X 'POST' \
  -H "Authorization: Bearer $TOKENHUB_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
  "model": "deepseek-v3.2-think",
  "input": "Explain why low latency matters for an AI product in one sentence."
}'

const response = await fetch("https://us-api.tokenhub.com/v1/responses", {
  method: "POST",
  headers: {
    Authorization: `Bearer ${process.env.TOKENHUB_API_KEY}`,
    "Content-Type": "application/json",
  },
  body: JSON.stringify({
    "model": "deepseek-v3.2-think",
    "input": "Explain why low latency matters for an AI product in one sentence."
  }),
})
const data = await response.json()
console.log(data)

import os
import requests

response = requests.request(method="POST", url="https://us-api.tokenhub.com/v1/responses",
    headers={
        "Authorization": f"Bearer {os.environ['TOKENHUB_API_KEY']}",
        "Content-Type": "application/json",
    },
    json=__import__("json").loads("{\"model\":\"deepseek-v3.2-think\",\"input\":\"Explain why low latency matters for an AI product in one sentence.\"}"),
)
response.raise_for_status()
print(response.json())

cURLTypeScript fetchPython requests

curl 'https://us-api.tokenhub.com/v1/messages' \
  -X 'POST' \
  -H "Authorization: Bearer $TOKENHUB_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
  "model": "deepseek-v3.2-think",
  "max_tokens": 1024,
  "messages": [
    {
      "role": "user",
      "content": "Explain why low latency matters for an AI product in one sentence."
    }
  ]
}'

const response = await fetch("https://us-api.tokenhub.com/v1/messages", {
  method: "POST",
  headers: {
    Authorization: `Bearer ${process.env.TOKENHUB_API_KEY}`,
    "Content-Type": "application/json",
  },
  body: JSON.stringify({
    "model": "deepseek-v3.2-think",
    "max_tokens": 1024,
    "messages": [
      {
        "role": "user",
        "content": "Explain why low latency matters for an AI product in one sentence."
      }
    ]
  }),
})
const data = await response.json()
console.log(data)

import os
import requests

response = requests.request(method="POST", url="https://us-api.tokenhub.com/v1/messages",
    headers={
        "Authorization": f"Bearer {os.environ['TOKENHUB_API_KEY']}",
        "Content-Type": "application/json",
    },
    json=__import__("json").loads("{\"model\":\"deepseek-v3.2-think\",\"max_tokens\":1024,\"messages\":[{\"role\":\"user\",\"content\":\"Explain why low latency matters for an AI product in one sentence.\"}]}"),
)
response.raise_for_status()
print(response.json())

應該選擇哪個 DeepSeek 模型？

依 DeepSeek API 的實際任務選擇模型；接入前再確認 DeepSeek API 價格、上下文視窗與端點支援情況。

API 任務	優先查看	適合情境	上線前確認
高頻、重視成本的 API 呼叫	DeepSeek V4 Flash	聊天、摘要、內容生成和批次處理等重視快速迭代的任務。	DeepSeek API 價格、吞吐量與輸出 token 成本
複雜推理	DeepSeek R1	多步驟分析、程式碼推理等更看重答案品質而非延遲的任務。	推理延遲、輸出長度與單次請求總成本
通用生產任務	DeepSeek V3.2 Think	助理、資訊擷取等透過一個 DeepSeek API 接入涵蓋的日常產品功能。	任務準確率、端點支援與生產環境價格
長輸入和長對話	DeepSeek V4 Flash	提示詞、文件或對話歷史成為限制的長上下文 DeepSeek API 請求。	上下文視窗、輸入 token 價格與串流輸出支援
高品質或高風險交付	DeepSeek V4 Pro	最終交付更看重能力上限，而不是追求最低單位成本的任務。	具體任務品質、延遲，以及與 Flash 的成本差異