deepseek-r1
DeepSeek
Open-source reasoning model with visible chain-of-thought, tuned for math, coding, and multi-step problem solving.
$0.595/ M
$2.125/ M
$0/ M
$0/ M
Faster, lower-cost variant of DeepSeek V4, optimized for coding assistants, high-throughput chat, and latency-sensitive Agent workflows.
Giá đầu vào
Thị trường: $0.1477
/ M
Giá đầu ra
Thị trường: $0.2955
/ M
Đọc cache
Thị trường: $0.0029
/ M
Ghi cache
/ M
Đầu vào
Văn bản
Đầu ra
Văn bản
{
"model": "deepseek/deepseek-v4-flash",
"messages": [
{
"role": "user",
"content": "你好!请先介绍一下你自己,包括你的主要能力与适用场景,并给一个简短的使用示例。"
}
],
"stream": false
}{}API
Mã sẵn sàng để gọi mô hình này qua API tương thích OpenAI của EasyAPI.
Tạo API token trong console và đặt làm biến môi trường:
Thay sk-xxxxxxxx bằng mã API từ console (Bearer Token).
Tạo API tokenexport EASYAPI_API_KEY=sk-xxxxxxxxDùng deepseek/deepseek-v4-flash với API EasyAPI:
curl https://token.easyapi.com/v1/chat/completions \
-H "Content-Type: application/json" \
-H "Authorization: Bearer $EASYAPI_API_KEY" \
-d '{"model":"deepseek/deepseek-v4-flash","messages":[{"role":"user","content":"Hello!"}]}'Thêm "stream": true vào body để nhận phản hồi qua SSE:
curl -N https://token.easyapi.com/v1/chat/completions \
-H "Content-Type: application/json" \
-H "Authorization: Bearer $EASYAPI_API_KEY" \
-d '{"model":"deepseek/deepseek-v4-flash","messages":[{"role":"user","content":"Hello!"}],"stream":true}'Gửi yêu cầu chat completion. Hỗ trợ streaming và non-streaming.
https://token.easyapi.com/v1/chat/completionsBearer $EASYAPI_API_KEYapplication/jsondeepseek/deepseek-v4-flash发送对话补全请求,支持流式与非流式。 Tài liệu
https://token.easyapi.com/v1/modelsBearer $EASYAPI_API_KEY获取当前账户可用模型列表。 Tài liệu
| Tên | Kiểu | Mặc định | Mô tả |
|---|---|---|---|
model | string | — | 模型 ID,如 openai/gpt-4o、anthropic/claude-3-5-sonnet |
messages | array | — | 对话消息列表,每项含 role 与 content |
stream | boolean | false | 设为 true 时以 SSE 流式返回 |
temperature | float | 1 | 采样温度,越高越随机 |
max_tokens | integer | — | 单次回复最大生成 Token 数 |
top_p | float | 1 | 核采样,限制候选词累积概率 |
frequency_penalty | float | 0 | 频率惩罚,降低重复用词 |
presence_penalty | float | 0 | 存在惩罚,鼓励新话题 |
SDK chính thức OpenAI và khác hoạt động khi đổi base URL. Xem hướng dẫn tích hợp
DeepSeek
Open-source reasoning model with visible chain-of-thought, tuned for math, coding, and multi-step problem solving.
$0.595/ M
$2.125/ M
$0/ M
$0/ M
DeepSeek
Open-source general-purpose LLM with representative instruction-following and coding skills within the open-weights ecosystem, suited for chat, code assistance, and enterprise text workflows.
$0.2439/ M
$0.975/ M
$0.0484/ M
$0/ M
DeepSeek
Hybrid model that switches between thinking and non-thinking modes on a single endpoint, post-trained for tool use and code agent workflows.
$0.4879/ M
$1.4629/ M
$0.1105/ M
$0/ M
DeepSeek
Best value for code and math. Fraction of flagship cost.
$0.2439/ M
$0.3663/ M
$0.0493/ M
$0.3051/ M