DeepSeek

可使用

deepseek-v4-flash-vision-exp

DeepSeek-V4-Flash-Vision API

Name: DeepSeek-V4-Flash-Vision API
Brand: CometAPI
SKU: deepseek-v4-flash-vision-exp
Price: 0.44 USD
Availability: InStock

deepseek-v4-flash-vision-exp 模型可同時接收圖片與文字，因此你可以讓模型描述圖片、讀取螢幕截圖中的文字、分析圖表等。支援的圖片格式：JPEG、PNG、GIF 與 WebP。格式是依據實際檔案內容偵測，而非依據檔名或宣告的 MIME 類型。

起始價格

$0.352 / 100 萬輸入 tokens

按使用量計費

API 健康狀態

可使用

正式環境端點

回應時間

支援非同步

實測 API 回應時間

發布時間

Aug 20, 2026

DeepSeek

Playground

透過實際請求試用 DeepSeek-V4-Flash-Vision,檢視參數並在整合 API 前驗證輸出結果。

定價

先參考上方的單次執行預估,再查看完整的 CometAPI 與官方價格比較。

第一次請求

—

預估 CometAPI 使用費用

十次請求

—

預估 CometAPI 使用費用

十五次請求

—

預估 CometAPI 使用費用

DeepSeek-V4-Flash-Vision 的定價

探索 DeepSeek-V4-Flash-Vision 的競爭性定價，專為滿足各種預算和使用需求而設計。我們靈活的方案確保您只需為實際使用量付費，讓您能夠隨著需求增長輕鬆擴展。了解 DeepSeek-V4-Flash-Vision 如何在保持成本可控的同時提升您的專案效果。

Model	Comet Price (USD / M Tokens)	Official Price (USD / M Tokens)	Discount
deepseek-v4-flash-vision	輸入:$0.352/M 輸出:$1.056/M	輸入:$0.44/M 輸出:$1.32/M	-20%
deepseek-v4-flash-vision-exp	輸入:$0.352/M 輸出:$1.056/M	輸入:$0.44/M 輸出:$1.32/M	-20%

API

複製可用的端點與程式碼範例,需要完整參數時再開啟完整 API 參考文件。

一個 API,直接對接正式環境模型

只需驗證一次,即可呼叫模型端點,並在各供應商間維持一致的計費與監控流程。

DeepSeek-V4-Flash-Vision 的範例程式碼和 API

存取完整的範例程式碼和 API 資源，以簡化您的 DeepSeek-V4-Flash-Vision 整合流程。我們詳盡的文件提供逐步指引，協助您在專案中充分發揮 DeepSeek-V4-Flash-Vision 的潛力。

#!/usr/bin/env bash

# Get your CometAPI key from https://www.cometapi.com/console/token
COMETAPI_KEY="${COMETAPI_KEY:-<YOUR_COMETAPI_KEY>}"
IMAGE_BASE64="${IMAGE_BASE64:-<BASE64_DATA>}"

chat_response=$(curl --silent --show-error --fail-with-body --max-time 180 --retry 0 \
  --request POST "https://api.cometapi.com/v1/chat/completions" \
  --header "Authorization: Bearer $COMETAPI_KEY" \
  --header "Content-Type: application/json" \
  --data-raw "$(jq -nc --arg model 'deepseek-v4-flash-vision-exp' --arg image_base64 "$IMAGE_BASE64" '{
    model: $model,
    thinking: {type: "disabled"},
    messages: [{
      role: "user",
      content: [
        {type: "text", text: "Describe the main visual elements in this image in one concise sentence."},
        {type: "image_url", image_url: {url: ("data:image/png;base64," + $image_base64)}}
      ]
    }],
    max_tokens: 512
  }')" \
  ${COMETAPI_RUNTIME_EVIDENCE:+--write-out $'%{stderr}__COMETAPI_RUNTIME__{"method":"POST","path":"/v1/chat/completions","http_status":%{http_code},"headers":{"content-type":"%header{content-type}","x-cometapi-request-id":"%header{x-cometapi-request-id}","x-request-id":"%header{x-request-id}","request-id":"%header{request-id}","cf-ray":"%header{cf-ray}"}}
'})

printf 'Chat Completions response:
'
printf '%s
' "$chat_response" | jq -r '.choices[0].message.content'

cURL Code Example

#!/usr/bin/env bash

# Get your CometAPI key from https://www.cometapi.com/console/token
COMETAPI_KEY="${COMETAPI_KEY:-<YOUR_COMETAPI_KEY>}"
IMAGE_BASE64="${IMAGE_BASE64:-<BASE64_DATA>}"

chat_response=$(curl --silent --show-error --fail-with-body --max-time 180 --retry 0 \
  --request POST "https://api.cometapi.com/v1/chat/completions" \
  --header "Authorization: Bearer $COMETAPI_KEY" \
  --header "Content-Type: application/json" \
  --data-raw "$(jq -nc --arg model 'deepseek-v4-flash-vision-exp' --arg image_base64 "$IMAGE_BASE64" '{
    model: $model,
    thinking: {type: "disabled"},
    messages: [{
      role: "user",
      content: [
        {type: "text", text: "Describe the main visual elements in this image in one concise sentence."},
        {type: "image_url", image_url: {url: ("data:image/png;base64," + $image_base64)}}
      ]
    }],
    max_tokens: 512
  }')" \
  ${COMETAPI_RUNTIME_EVIDENCE:+--write-out $'%{stderr}__COMETAPI_RUNTIME__{"method":"POST","path":"/v1/chat/completions","http_status":%{http_code},"headers":{"content-type":"%header{content-type}","x-cometapi-request-id":"%header{x-cometapi-request-id}","x-request-id":"%header{x-request-id}","request-id":"%header{request-id}","cf-ray":"%header{cf-ray}"}}\n'})

printf 'Chat Completions response:\n'
printf '%s\n' "$chat_response" | jq -r '.choices[0].message.content'

Python Code Example

import json
import os
import sys

import requests


# Get your CometAPI key from https://www.cometapi.com/console/token
COMETAPI_KEY = os.environ.get("COMETAPI_KEY") or "<YOUR_COMETAPI_KEY>"
IMAGE_BASE64 = os.environ.get("IMAGE_BASE64") or "<BASE64_DATA>"
IMAGE_DATA_URL = f"data:image/png;base64,{IMAGE_BASE64}"
MODEL = "deepseek-v4-flash-vision-exp"
PROMPT = "Describe the main visual elements in this image in one concise sentence."
HEADERS = {
    "Authorization": f"Bearer {COMETAPI_KEY}",
    "Content-Type": "application/json",
}


def capture_runtime(response, path):
    if os.environ.get("COMETAPI_RUNTIME_EVIDENCE") != "1":
        return
    safe_headers = {
        name: response.headers.get(name)
        for name in (
            "cf-ray",
            "content-length",
            "content-type",
            "date",
            "request-id",
            "server",
            "x-cometapi-request-id",
            "x-request-id",
        )
        if response.headers.get(name) is not None
    }
    print(
        "__COMETAPI_RUNTIME__"
        + json.dumps(
            {
                "method": "POST",
                "path": path,
                "http_status": response.status_code,
                "headers": safe_headers,
            }
        ),
        file=sys.stderr,
    )


chat_response = requests.post(
    "https://api.cometapi.com/v1/chat/completions",
    headers=HEADERS,
    json={
        "model": MODEL,
        "thinking": {"type": "disabled"},
        "messages": [
            {
                "role": "user",
                "content": [
                    {"type": "text", "text": PROMPT},
                    {"type": "image_url", "image_url": {"url": IMAGE_DATA_URL}},
                ],
            }
        ],
        "max_tokens": 512,
    },
    timeout=180,
)
capture_runtime(chat_response, "/v1/chat/completions")
chat_response.raise_for_status()
chat = chat_response.json()
print("Chat Completions response:")
print(chat["choices"][0]["message"]["content"])

JavaScript Code Example

// Get your CometAPI key from https://www.cometapi.com/console/token
const apiKey = process.env.COMETAPI_KEY || "<YOUR_COMETAPI_KEY>";
const imageBase64 = process.env.IMAGE_BASE64 || "<BASE64_DATA>";
const imageDataUrl = `data:image/png;base64,${imageBase64}`;
const model = "deepseek-v4-flash-vision-exp";
const prompt = "Describe the main visual elements in this image in one concise sentence.";
const headers = {
  Authorization: `Bearer ${apiKey}`,
  "Content-Type": "application/json",
};

function captureRuntime(response, path) {
  if (process.env.COMETAPI_RUNTIME_EVIDENCE !== "1") return;
  const safeHeaderNames = [
    "cf-ray",
    "content-length",
    "content-type",
    "date",
    "request-id",
    "server",
    "x-cometapi-request-id",
    "x-request-id",
  ];
  const safeHeaders = Object.fromEntries(
    safeHeaderNames
      .map((name) => [name, response.headers.get(name)])
      .filter(([, value]) => value !== null),
  );
  console.error(
    `__COMETAPI_RUNTIME__${JSON.stringify({
      method: "POST",
      path,
      http_status: response.status,
      headers: safeHeaders,
    })}`,
  );
}

const chatResponse = await fetch("https://api.cometapi.com/v1/chat/completions", {
  method: "POST",
  headers,
  body: JSON.stringify({
    model,
    thinking: { type: "disabled" },
    messages: [
      {
        role: "user",
        content: [
          { type: "text", text: prompt },
          { type: "image_url", image_url: { url: imageDataUrl } },
        ],
      },
    ],
    max_tokens: 512,
  }),
  signal: AbortSignal.timeout(180_000),
});
captureRuntime(chatResponse, "/v1/chat/completions");
if (!chatResponse.ok) throw new Error(`Chat Completions failed: HTTP ${chatResponse.status}`);
const chat = await chatResponse.json();
console.log("Chat Completions response:");
console.log(chat.choices[0].message.content);

功能

在選擇架構或評估正式環境負載前,先掌握模型的關鍵資訊。

支援的功能

現有模型 API 回傳的功能。

text-to-text

image-to-text

輸入與輸出

輸入

text

輸出

text

正式環境資訊

供應商: DeepSeek
上下文: 依模型而定
商業使用: 支援

應用場景

將 DeepSeek-V4-Flash-Vision 用於符合其 text 能力的正式工作流程,並在長期整合前比較其他選項。

正式環境聊天與 Agent 工作流程

程式撰寫、分析與結構化生成

透過單一 API 進行大量自動化

其他模型

比較 CometAPI 提供的其他模型,在品質、延遲、功能與價格間取得平衡。

開啟模型比較 →

Google

Gemini 3.7 Flash

3.7 Flash 在軟體工程、知識工作與網頁開發的工作流程中帶來顯著提升 — 並以 3.6 Flash 原本每百萬個 Token 成本的一半作為導入價格。

查看模型

DeepSeek

DeepSeek V4 Pro

DeepSeek V4 Pro 是 DeepSeek 推出的大規模混合專家模型，總參數量達 1.6T，啟用參數為 49B，並支援 1M-token 的上下文視窗。它專為高階推理、程式設計與長時域智能體工作流程而設計，並在知識、數學與軟體工程等基準測試上表現強勁。

查看模型

xAI

Grok 4.6

面向程式碼及其他一切的旗艦模型：代理式工具呼叫、極少幻覺、可設定的推理能力。

查看模型

Zhipu AI

GLM 5.3

GLM-5.3 是新一代開放原始碼大型語言模型，針對複雜程式設計、長時程任務與資安情境進行了優化。相較於 GLM-5.2，它在程式設計能力與智能體表現方面有顯著提升。

查看模型

Anthropic

Claude Fable 5

對 Claude Fable 5 的存取權限已恢復。它為您最具雄心的程式設計與專業工作帶來第5代智慧。

用於程式設計與智能體的高效能模型

Uptime

在正式上線前,查看即時心跳資料、端點可用性與實測回應時間。

版本

在鎖定正式環境整合版本前,先查看可用的模型識別碼。

deepseek-v4-flash-vision

deepseek-v4-flash-vision-exp

持續探索

比較模型,繼續你的研究

DeepSeek-V4-Flash-Vision 的定價

Model	Comet Price (USD / M Tokens)	Official Price (USD / M Tokens)	Discount
deepseek-v4-flash-vision	輸入:$0.352/M 輸出:$1.056/M	輸入:$0.44/M 輸出:$1.32/M	-20%
deepseek-v4-flash-vision-exp	輸入:$0.352/M 輸出:$1.056/M	輸入:$0.44/M 輸出:$1.32/M	-20%

DeepSeek-V4-Flash-Vision 的範例程式碼和 API

#!/usr/bin/env bash

# Get your CometAPI key from https://www.cometapi.com/console/token
COMETAPI_KEY="${COMETAPI_KEY:-<YOUR_COMETAPI_KEY>}"
IMAGE_BASE64="${IMAGE_BASE64:-<BASE64_DATA>}"

chat_response=$(curl --silent --show-error --fail-with-body --max-time 180 --retry 0 \
  --request POST "https://api.cometapi.com/v1/chat/completions" \
  --header "Authorization: Bearer $COMETAPI_KEY" \
  --header "Content-Type: application/json" \
  --data-raw "$(jq -nc --arg model 'deepseek-v4-flash-vision-exp' --arg image_base64 "$IMAGE_BASE64" '{
    model: $model,
    thinking: {type: "disabled"},
    messages: [{
      role: "user",
      content: [
        {type: "text", text: "Describe the main visual elements in this image in one concise sentence."},
        {type: "image_url", image_url: {url: ("data:image/png;base64," + $image_base64)}}
      ]
    }],
    max_tokens: 512
  }')" \
  ${COMETAPI_RUNTIME_EVIDENCE:+--write-out $'%{stderr}__COMETAPI_RUNTIME__{"method":"POST","path":"/v1/chat/completions","http_status":%{http_code},"headers":{"content-type":"%header{content-type}","x-cometapi-request-id":"%header{x-cometapi-request-id}","x-request-id":"%header{x-request-id}","request-id":"%header{request-id}","cf-ray":"%header{cf-ray}"}}
'})

printf 'Chat Completions response:
'
printf '%s
' "$chat_response" | jq -r '.choices[0].message.content'

cURL Code Example

#!/usr/bin/env bash

# Get your CometAPI key from https://www.cometapi.com/console/token
COMETAPI_KEY="${COMETAPI_KEY:-<YOUR_COMETAPI_KEY>}"
IMAGE_BASE64="${IMAGE_BASE64:-<BASE64_DATA>}"

chat_response=$(curl --silent --show-error --fail-with-body --max-time 180 --retry 0 \
  --request POST "https://api.cometapi.com/v1/chat/completions" \
  --header "Authorization: Bearer $COMETAPI_KEY" \
  --header "Content-Type: application/json" \
  --data-raw "$(jq -nc --arg model 'deepseek-v4-flash-vision-exp' --arg image_base64 "$IMAGE_BASE64" '{
    model: $model,
    thinking: {type: "disabled"},
    messages: [{
      role: "user",
      content: [
        {type: "text", text: "Describe the main visual elements in this image in one concise sentence."},
        {type: "image_url", image_url: {url: ("data:image/png;base64," + $image_base64)}}
      ]
    }],
    max_tokens: 512
  }')" \
  ${COMETAPI_RUNTIME_EVIDENCE:+--write-out $'%{stderr}__COMETAPI_RUNTIME__{"method":"POST","path":"/v1/chat/completions","http_status":%{http_code},"headers":{"content-type":"%header{content-type}","x-cometapi-request-id":"%header{x-cometapi-request-id}","x-request-id":"%header{x-request-id}","request-id":"%header{request-id}","cf-ray":"%header{cf-ray}"}}\n'})

printf 'Chat Completions response:\n'
printf '%s\n' "$chat_response" | jq -r '.choices[0].message.content'

Python Code Example

import json
import os
import sys

import requests


# Get your CometAPI key from https://www.cometapi.com/console/token
COMETAPI_KEY = os.environ.get("COMETAPI_KEY") or "<YOUR_COMETAPI_KEY>"
IMAGE_BASE64 = os.environ.get("IMAGE_BASE64") or "<BASE64_DATA>"
IMAGE_DATA_URL = f"data:image/png;base64,{IMAGE_BASE64}"
MODEL = "deepseek-v4-flash-vision-exp"
PROMPT = "Describe the main visual elements in this image in one concise sentence."
HEADERS = {
    "Authorization": f"Bearer {COMETAPI_KEY}",
    "Content-Type": "application/json",
}


def capture_runtime(response, path):
    if os.environ.get("COMETAPI_RUNTIME_EVIDENCE") != "1":
        return
    safe_headers = {
        name: response.headers.get(name)
        for name in (
            "cf-ray",
            "content-length",
            "content-type",
            "date",
            "request-id",
            "server",
            "x-cometapi-request-id",
            "x-request-id",
        )
        if response.headers.get(name) is not None
    }
    print(
        "__COMETAPI_RUNTIME__"
        + json.dumps(
            {
                "method": "POST",
                "path": path,
                "http_status": response.status_code,
                "headers": safe_headers,
            }
        ),
        file=sys.stderr,
    )


chat_response = requests.post(
    "https://api.cometapi.com/v1/chat/completions",
    headers=HEADERS,
    json={
        "model": MODEL,
        "thinking": {"type": "disabled"},
        "messages": [
            {
                "role": "user",
                "content": [
                    {"type": "text", "text": PROMPT},
                    {"type": "image_url", "image_url": {"url": IMAGE_DATA_URL}},
                ],
            }
        ],
        "max_tokens": 512,
    },
    timeout=180,
)
capture_runtime(chat_response, "/v1/chat/completions")
chat_response.raise_for_status()
chat = chat_response.json()
print("Chat Completions response:")
print(chat["choices"][0]["message"]["content"])

JavaScript Code Example

// Get your CometAPI key from https://www.cometapi.com/console/token
const apiKey = process.env.COMETAPI_KEY || "<YOUR_COMETAPI_KEY>";
const imageBase64 = process.env.IMAGE_BASE64 || "<BASE64_DATA>";
const imageDataUrl = `data:image/png;base64,${imageBase64}`;
const model = "deepseek-v4-flash-vision-exp";
const prompt = "Describe the main visual elements in this image in one concise sentence.";
const headers = {
  Authorization: `Bearer ${apiKey}`,
  "Content-Type": "application/json",
};

function captureRuntime(response, path) {
  if (process.env.COMETAPI_RUNTIME_EVIDENCE !== "1") return;
  const safeHeaderNames = [
    "cf-ray",
    "content-length",
    "content-type",
    "date",
    "request-id",
    "server",
    "x-cometapi-request-id",
    "x-request-id",
  ];
  const safeHeaders = Object.fromEntries(
    safeHeaderNames
      .map((name) => [name, response.headers.get(name)])
      .filter(([, value]) => value !== null),
  );
  console.error(
    `__COMETAPI_RUNTIME__${JSON.stringify({
      method: "POST",
      path,
      http_status: response.status,
      headers: safeHeaders,
    })}`,
  );
}

const chatResponse = await fetch("https://api.cometapi.com/v1/chat/completions", {
  method: "POST",
  headers,
  body: JSON.stringify({
    model,
    thinking: { type: "disabled" },
    messages: [
      {
        role: "user",
        content: [
          { type: "text", text: prompt },
          { type: "image_url", image_url: { url: imageDataUrl } },
        ],
      },
    ],
    max_tokens: 512,
  }),
  signal: AbortSignal.timeout(180_000),
});
captureRuntime(chatResponse, "/v1/chat/completions");
if (!chatResponse.ok) throw new Error(`Chat Completions failed: HTTP ${chatResponse.status}`);
const chat = await chatResponse.json();
console.log("Chat Completions response:");
console.log(chat.choices[0].message.content);