支持的模型

接入 Claude、GPT、o-series 等主流 AI 模型，一个 API Key 搞定一切

Anthropic

Claude family — known for helpfulness, harmlessness, and honesty.

Claude Opus 4.6

claude-opus-4-6

可用

Most capable Claude model with deep reasoning, advanced analysis, and creative generation. Ideal for complex tasks requiring nuanced understanding.

输入价格

$15/M tok

输出价格

$75/M tok

上下文 200,000 tokens|最大输出 16,384

对话视觉代码推理流式函数调用

Claude Sonnet 4.6

claude-sonnet-4-6

可用

Best balance of speed and capability. Excellent for most production workloads including coding, analysis, and content generation.

输入价格

$3/M tok

输出价格

$15/M tok

上下文 200,000 tokens|最大输出 16,384

对话视觉代码推理流式函数调用

Claude Haiku 4.5

claude-haiku-4-5

可用

Fastest and most affordable Claude model. Perfect for high-throughput tasks like classification, extraction, and simple Q&A.

输入价格

$0.8/M tok

输出价格

$4/M tok

上下文 200,000 tokens|最大输出 8,192

对话视觉代码流式函数调用

OpenAI

GPT and o-series — industry-standard models with broad capabilities.

GPT-5.4

gpt-5-4

可用

Latest and most powerful GPT model with advanced reasoning, multimodal understanding, and superior instruction following.

输入价格

$10/M tok

输出价格

$30/M tok

上下文 128,000 tokens|最大输出 16,384

对话视觉代码推理流式函数调用JSON模式

GPT-4o

gpt-4o

可用

Efficient multimodal model with strong capabilities across text and vision. Great balance of quality and speed.

输入价格

$2.5/M tok

输出价格

$10/M tok

上下文 128,000 tokens|最大输出 16,384

对话视觉代码流式函数调用JSON模式

o3

o3

可用

Advanced reasoning model that excels at math, science, coding, and complex logical analysis. Uses extended thinking for deeper problem-solving.

输入价格

$10/M tok

输出价格

$40/M tok

上下文 200,000 tokens|最大输出 100,000

对话代码推理流式函数调用JSON模式

o3-pro

o3-pro

可用

Most powerful reasoning model with extended thinking. Designed for the hardest problems in math, science, and engineering.

输入价格

$20/M tok

输出价格

$80/M tok

上下文 200,000 tokens|最大输出 100,000

对话代码推理流式函数调用JSON模式

o4-mini

o4-mini

可用

Affordable reasoning model with strong performance for everyday analytical tasks. Best cost-efficiency in the o-series.

输入价格

$1.5/M tok

输出价格

$6/M tok

上下文 200,000 tokens|最大输出 100,000

对话代码推理流式函数调用JSON模式

Codex

codex

可用

Specialized coding model optimized for code generation, editing, and understanding. Supports a wide range of programming languages.

输入价格

$3/M tok

输出价格

$12/M tok

上下文 128,000 tokens|最大输出 16,384

对话代码流式函数调用

Google

Gemini family — Google's most capable multimodal models with massive context windows.

Gemini 2.5 Pro

gemini-2.5-pro

可用

Most capable Gemini model with deep thinking, advanced reasoning, and multimodal understanding. Excels at complex analysis and coding tasks.

输入价格

$1.25/M tok

输出价格

$10/M tok

上下文 1,048,576 tokens|最大输出 65,536

对话视觉代码推理流式函数调用JSON模式thinking

Gemini 2.5 Flash

gemini-2.5-flash

可用

Fast and cost-efficient Gemini model with thinking capabilities. Great balance of speed, quality, and price for production workloads.

输入价格

$0.15/M tok

输出价格

$0.6/M tok

上下文 1,048,576 tokens|最大输出 65,536

对话视觉代码推理流式函数调用JSON模式thinking

Gemini 2.0 Flash

gemini-2.0-flash

可用

Fast Gemini model optimized for low-latency applications. Ideal for real-time chat, classification, and quick Q&A tasks.

输入价格

$0.1/M tok

输出价格

$0.4/M tok

上下文 1,048,576 tokens|最大输出 8,192

对话视觉代码流式函数调用JSON模式

兼容格式

DeepNorth AI API 完全兼容 OpenAI API 格式，无需修改代码即可无缝迁移

OpenAI SDK 兼容

直接使用 OpenAI Python/Node.js SDK，只需修改 base_url 和 api_key。

# Python 示例

from openai import OpenAI

client = OpenAI(

base_url=https://api.deepnorth.ai/v1,

api_key=your-key

)

HTTP 请求兼容

任何支持 HTTP 请求的工具和框架都可以直接对接，标准 RESTful API。

# cURL 示例

curl https://api.deepnorth.ai/v1/chat/completions

-H 'Authorization: Bearer your-key'

-H 'Content-Type: application/json'

-d '{"model": "claude-sonnet-4-6", ...}'

流式输出兼容

完全支持 SSE 流式传输，与 OpenAI 流式响应格式完全一致。

# 流式请求

response = client.chat.completions.create(

model="claude-sonnet-4-6",

messages=[...],

stream=True

)

DeepNorth

支持的模型

接入 Claude、GPT、o-series 等主流 AI 模型，一个 API Key 搞定一切

Anthropic

Claude family — known for helpfulness, harmlessness, and honesty.

Claude Opus 4.6

claude-opus-4-6

可用

Most capable Claude model with deep reasoning, advanced analysis, and creative generation. Ideal for complex tasks requiring nuanced understanding.

输入价格

$15/M tok

输出价格

$75/M tok

上下文 200,000 tokens|最大输出 16,384

对话视觉代码推理流式函数调用

Claude Sonnet 4.6

claude-sonnet-4-6

可用

Best balance of speed and capability. Excellent for most production workloads including coding, analysis, and content generation.

输入价格

$3/M tok

输出价格

$15/M tok

上下文 200,000 tokens|最大输出 16,384

对话视觉代码推理流式函数调用

Claude Haiku 4.5

claude-haiku-4-5

可用

Fastest and most affordable Claude model. Perfect for high-throughput tasks like classification, extraction, and simple Q&A.

输入价格

$0.8/M tok

输出价格

$4/M tok

上下文 200,000 tokens|最大输出 8,192

对话视觉代码流式函数调用

OpenAI

GPT and o-series — industry-standard models with broad capabilities.

GPT-5.4

gpt-5-4

可用

Latest and most powerful GPT model with advanced reasoning, multimodal understanding, and superior instruction following.

输入价格

$10/M tok

输出价格

$30/M tok

上下文 128,000 tokens|最大输出 16,384

对话视觉代码推理流式函数调用JSON模式

GPT-4o

gpt-4o

可用

Efficient multimodal model with strong capabilities across text and vision. Great balance of quality and speed.

输入价格

$2.5/M tok

输出价格

$10/M tok

上下文 128,000 tokens|最大输出 16,384

对话视觉代码流式函数调用JSON模式

o3

o3

可用

Advanced reasoning model that excels at math, science, coding, and complex logical analysis. Uses extended thinking for deeper problem-solving.

输入价格

$10/M tok

输出价格

$40/M tok

上下文 200,000 tokens|最大输出 100,000

对话代码推理流式函数调用JSON模式

o3-pro

o3-pro

可用

Most powerful reasoning model with extended thinking. Designed for the hardest problems in math, science, and engineering.

输入价格

$20/M tok

输出价格

$80/M tok

上下文 200,000 tokens|最大输出 100,000

对话代码推理流式函数调用JSON模式

o4-mini

o4-mini

可用

Affordable reasoning model with strong performance for everyday analytical tasks. Best cost-efficiency in the o-series.

输入价格

$1.5/M tok

输出价格

$6/M tok

上下文 200,000 tokens|最大输出 100,000

对话代码推理流式函数调用JSON模式

Codex

codex

可用

Specialized coding model optimized for code generation, editing, and understanding. Supports a wide range of programming languages.

输入价格

$3/M tok

输出价格

$12/M tok

上下文 128,000 tokens|最大输出 16,384

对话代码流式函数调用

Google

Gemini family — Google's most capable multimodal models with massive context windows.

Gemini 2.5 Pro

gemini-2.5-pro

可用

Most capable Gemini model with deep thinking, advanced reasoning, and multimodal understanding. Excels at complex analysis and coding tasks.

输入价格

$1.25/M tok

输出价格

$10/M tok

上下文 1,048,576 tokens|最大输出 65,536

对话视觉代码推理流式函数调用JSON模式thinking

Gemini 2.5 Flash

gemini-2.5-flash

可用

Fast and cost-efficient Gemini model with thinking capabilities. Great balance of speed, quality, and price for production workloads.

输入价格

$0.15/M tok

输出价格

$0.6/M tok

上下文 1,048,576 tokens|最大输出 65,536

对话视觉代码推理流式函数调用JSON模式thinking

Gemini 2.0 Flash

gemini-2.0-flash

可用

Fast Gemini model optimized for low-latency applications. Ideal for real-time chat, classification, and quick Q&A tasks.

输入价格

$0.1/M tok

输出价格

$0.4/M tok

上下文 1,048,576 tokens|最大输出 8,192

对话视觉代码流式函数调用JSON模式

兼容格式

DeepNorth AI API 完全兼容 OpenAI API 格式，无需修改代码即可无缝迁移

OpenAI SDK 兼容

直接使用 OpenAI Python/Node.js SDK，只需修改 base_url 和 api_key。

# Python 示例

from openai import OpenAI

client = OpenAI(

base_url=https://api.deepnorth.ai/v1,

api_key=your-key

)

HTTP 请求兼容

任何支持 HTTP 请求的工具和框架都可以直接对接，标准 RESTful API。

# cURL 示例

curl https://api.deepnorth.ai/v1/chat/completions

-H 'Authorization: Bearer your-key'

-H 'Content-Type: application/json'

-d '{"model": "claude-sonnet-4-6", ...}'

流式输出兼容

完全支持 SSE 流式传输，与 OpenAI 流式响应格式完全一致。

# 流式请求

response = client.chat.completions.create(

model="claude-sonnet-4-6",

messages=[...],

stream=True

)