使用文档 - Dengta AI

⚡

🔗

🛠

1

2

3

curl https://aiapi.dengta-learning.online/v1/chat/completions \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer sk-your-token-here" \
  -d '{
    "model": "gpt-5.4",
    "messages": [{"role": "user", "content": "Hello!"}]
  }'

📦

bash <(curl -fsSL https://ai.dengta-learning.online/install.sh)

irm https://ai.dengta-learning.online/install.ps1 -OutFile $env:TEMP\dti.ps1; & $env:TEMP\dti.ps1

🚀

bash <(curl -fsSL https://ai.dengta-learning.online/setup.sh)

irm https://ai.dengta-learning.online/setup.ps1 -OutFile $env:TEMP\dt.ps1; & $env:TEMP\dt.ps1

🐾

bash <(curl -fsSL https://ai.dengta-learning.online/openclaw-setup.sh)

irm https://ai.dengta-learning.online/openclaw-setup.ps1 -OutFile $env:TEMP\oc.ps1; & $env:TEMP\oc.ps1

DENGTA_KEY=sk-你的Key bash <(curl -fsSL https://ai.dengta-learning.online/openclaw-setup.sh)

1

2

3

1

2

3

1

2

bash

# 1. 全局安装
npm install -g @openai/codex

# 2. 首次启动（会自动生成 .codex 配置目录）
codex

toml

disable_response_storage = true
model = "gpt-5.4"
model_provider = "DengtaAI"
model_reasoning_effort = "high"

[model_providers."DengtaAI"]
name = "DengtaAI"
base_url = "https://aiapi.dengta-learning.online/v1"
requires_openai_auth = true
wire_api = "responses"

json

{
  "OPENAI_API_KEY": "sk-your-token-here"
}

bash

# 直接启动
codex

# 带一条指令启动
codex "帮我分析当前项目结构"

# 非交互执行
codex exec "检查当前仓库中有哪些 TODO"

json

{
  "env": {
    "ANTHROPIC_AUTH_TOKEN": "sk-your-token-here",
    "ANTHROPIC_BASE_URL": "https://aiapi.dengta-learning.online",
    "ANTHROPIC_DEFAULT_HAIKU_MODEL": "gpt-5.4",
    "ANTHROPIC_DEFAULT_OPUS_MODEL": "gpt-5.4",
    "ANTHROPIC_DEFAULT_SONNET_MODEL": "gpt-5.4",
    "ANTHROPIC_MODEL": "gpt-5.4",
    "ANTHROPIC_REASONING_MODEL": "gpt-5.4"
  }
}

bash

claude

bash

# 1. 全局安装
npm install -g openclaw@latest

# 2. 执行初始化引导
openclaw onboard

{
  "models": {
    "providers": {
      "dengta-ai": {
        "baseUrl": "https://aiapi.dengta-learning.online/v1",
        "apiKey": "sk-your-token-here",
        "api": "openai-completions",
        "headers": {
          "User-Agent": "Mozilla/5.0",
          "Accept": "application/json"
        },
        "models": [
          {
            "id": "gpt-5.4",
            "name": "GPT-5.4",
            "reasoning": true,
            "input": ["text", "image"],
            "cost": { "input": 0, "output": 0, "cacheRead": 0, "cacheWrite": 0 },
            "contextWindow": 200000,
            "maxTokens": 32768
          }
        ]
      }
    }
  },
  "agents": {
    "defaults": {
      "model": {
        "primary": "dengta-ai/gpt-5.4"
      }
    }
  }
}

bash

# 启动 OpenClaw
openclaw

# 如果需要启动 Gateway
openclaw gateway --port 18789

# 访问控制台：http://127.0.0.1:18789/

1

2

3

bash

# 全局安装
npm install -g opencode-ai

# 首次启动
opencode

json

{
  "$schema": "https://opencode.ai/config.json",
  "provider": {
    "DengtaAI": {
      "npm": "@ai-sdk/openai-compatible",
      "name": "DengtaAI",
      "options": {
        "baseURL": "https://aiapi.dengta-learning.online/v1"
      },
      "models": {
        "gpt-5.4": {
          "model": "gpt-5.4",
          "name": "GPT-5.4",
          "options": { "variant": "xhigh" },
          "limit": {
            "context": 200000,
            "output": 8192
          }
        }
      }
    }
  },
  "model": "DengtaAI/gpt-5.4",
  "small_model": "DengtaAI/gpt-5.4"
}

bash

# 重新启动
opencode

# 在 TUI 中查看模型列表
/models

# 或在终端直接检查
opencode models DengtaAI

1

2

3

4

text

https://aiapi.dengta-learning.online/v1


https://aiapi.dengta-learning.online/v1
https://ai.dengta-learning.online

http header

Authorization: Bearer sk-your-token-here


	GET /v1/models
	POST /v1/chat/completions
Responses	POST /v1/responses

curl https://aiapi.dengta-learning.online/v1/models \
  -H "Authorization: Bearer sk-your-token-here"


gpt-5.4
gpt-5.2
gpt-4o
gpt-4o-mini

Chat Completions

POST /v1/chat/completions


model	string
messages	array
stream	boolean
temperature	number
max_tokens	integer
tools	array
reasoning_effort	string

bash

curl https://aiapi.dengta-learning.online/v1/chat/completions \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer sk-your-token" \
  -d '{
    "model": "gpt-5.4",
    "messages": [
      {"role": "system", "content": "You are a helpful assistant."},
      {"role": "user", "content": "你好"}
    ],
    "stream": false
  }'

json

{
  "id": "chatcmpl-abc123",
  "object": "chat.completion",
  "created": 1709000000,
  "model": "gpt-5.4",
  "choices": [{
    "index": 0,
    "message": {
      "role": "assistant",
      "content": "你好！有什么我可以帮你的吗？"
    },
    "finish_reason": "stop"
  }],
  "usage": {
    "prompt_tokens": 20,
    "completion_tokens": 12,
    "total_tokens": 32
  }
}

Responses

POST /v1/responses


model	string
input	string/array
instructions	string
max_output_tokens	integer
reasoning.effort	string
stream	boolean

bash

curl https://aiapi.dengta-learning.online/v1/responses \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer sk-your-token-here" \
  -d '{
    "model": "gpt-5.4",
    "input": "请总结这段代码的作用，并列出三个风险点。",
    "reasoning": {"effort": "medium"},
    "max_output_tokens": 800
  }'

python

from openai import OpenAI

client = OpenAI(
    api_key="sk-your-token-here",
    base_url="https://aiapi.dengta-learning.online/v1"
)

response = client.chat.completions.create(
    model="gpt-5.4",
    messages=[{"role": "user", "content": "Hello!"}]
)
print(response.choices[0].message.content)

javascript

import OpenAI from 'openai';

const client = new OpenAI({
  apiKey: 'sk-your-token-here',
  baseURL: 'https://aiapi.dengta-learning.online/v1',
});

const response = await client.chat.completions.create({
  model: 'gpt-5.4',
  messages: [{ role: 'user', content: 'Hello!' }],
});
console.log(response.choices[0].message.content);

bash

# 查询可用模型
curl https://aiapi.dengta-learning.online/v1/models \
  -H "Authorization: Bearer sk-your-token-here"

# 发送对话请求
curl https://aiapi.dengta-learning.online/v1/chat/completions \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer sk-your-token-here" \
  -d '{"model":"gpt-5.4","messages":[{"role":"user","content":"Hello!"}]}'


400
401
403
404
429
500
502

code
invalid_api_key
insufficient_quota
model_not_found
context_length_exceeded