Quick Start

Welcome to AGIPower! This guide walks you through everything from signing up to making your first API request, step by step.

Just four simple steps to start using AGIPower:

Sign in to AGIPower: Visit the AGIPower login page and sign in using any of the following:
- Email
- GitHub account
- Google account
Choose a usage plan: Based on your use case, select Pay As You Go or Builder Plan.
Get an API key: After signing in, create an API key from the corresponding management page.
Make your first request: Pick the API protocol you’re most familiar with, copy the code example below, replace your API key, and run it.

Choose a Usage Plan

AGIPower offers two usage plans to fit different scenarios:

Pay As You Go

Best for: Production environments, commercial products, high-concurrency applications Pay As You Go uses a prepaid balance + usage-based billing model, purpose-built for production workloads:

No rate limits — supports high-concurrency calls
Per-token billing — transparent and predictable costs

Go to the Pay As You Go management page to top up your balance and create an API key.

For complete instructions, see the Pay As You Go Guide.

Builder Plan (Subscription)

Best for: Personal development, learning and exploration, Vibe Coding, rapid prototyping Builder Plan provides a fixed monthly fee with predictable AI model access:

Fixed monthly fee starting at $20/month
All-star model lineup — one subscription unlocks 100+ top models worldwide
Full-scenario coverage — Coding + image generation + video generation + chat, all in one place

Go to the Pricing page to view plan details and subscribe. Subscription plan comparison

For complete instructions, see the Subscription Plans Guide.

Plan Comparison

Dimension	Pay As You Go	Builder Plan (Subscription)
Best for	Production, commercial products	Personal development, learning
Billing	Usage-based, per-token	Fixed monthly fee
Rate Limit	Unlimited	10–15 RPM
Concurrency	Unlimited	Weekly Limit applies
Restrictions	None	Not allowed for production use

If your project is already live or heading toward commercialization, you must use the Pay As You Go plan. The subscription plan is strictly for personal development and learning — production use is prohibited.

Each model on the AGIPower platform has a unique slug. You can find model slugs on the Models page: model-slug

Or on a model’s detail page: model-slug

Supported API Protocols

AGIPower supports four major API protocols, letting you use your preferred SDK to call any model on the platform:

Protocol	Base URL	Compatible SDK	Description
OpenAI Chat Completions	`https://api.agipower.ai/v1`	OpenAI SDK	The most widely used Chat API
OpenAI Responses	`https://api.agipower.ai/v1`	OpenAI SDK	OpenAI’s next-gen Responses API
Anthropic Messages	`https://agipower.ai`	Anthropic SDK	Native protocol for the Claude family
Google Gemini	`https://api.agipower.ai`	Google GenAI SDK	Native protocol for the Gemini family

One of AGIPower’s core strengths is protocol agnosticism — you can call any model through any supported protocol. For example, use the OpenAI SDK to call Claude models, or the Anthropic SDK to call Gemini models.

Protocol 1: OpenAI Chat Completions

AGIPower’s API endpoint is fully compatible with the OpenAI Chat Completions API. Just change base_url and api_key to switch seamlessly.

Python

from openai import OpenAI

# 1. Initialize the OpenAI client
client = OpenAI(
    # 2. Point the base URL to the AGIPower endpoint
    base_url="https://api.agipower.ai/v1", 
    # 3. Replace with the API key from your AGIPower console
    api_key="<your AGIPower_API_KEY>", 
)

# 4. Create a Chat Completion request
completion = client.chat.completions.create(
    # 5. Specify the model in "provider/model-name" format
    model="google/gemini-3.1-pro-preview", 
    messages=[
        {
            "role": "user",
            "content": "What is the meaning of life?"
        }
    ]
)

print(completion.choices[0].message.content)

TypeScript

import OpenAI from "openai";

// 1. Initialize the OpenAI client
const openai = new OpenAI({
  // 2. Point the base URL to the AGIPower endpoint
  baseURL: "https://api.agipower.ai/v1", 
  // 3. Replace with the API key from your AGIPower console
  apiKey: "<your AGIPower_API_KEY>", 
});

async function main() {
  // 4. Create a Chat Completion request
  const completion = await openai.chat.completions.create({
    // 5. Specify the model in "provider/model-name" format
    model: "google/gemini-3.1-pro-preview", 
    messages: [
      {
        role: "user",
        content: "What is the meaning of life?",
      },
    ],
  });

  console.log(completion.choices[0].message.content);
}

main();

cURL

curl https://api.agipower.ai/v1/chat/completions \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer $AGIPower_API_KEY" \
  -d '{
    "model": "google/gemini-3.1-pro-preview",
    "messages": [
      {
        "role": "user",
        "content": "What is the meaning of life?"
      }
    ]
  }'

Protocol 2: OpenAI Responses

AGIPower is also compatible with OpenAI’s next-gen Responses API. The base URL is the same as Chat Completions, with a more streamlined calling convention.

Python

from openai import OpenAI

client = OpenAI(
    base_url="https://api.agipower.ai/v1", 
    api_key="<your AGIPower_API_KEY>", 
)

# Create a request using the Responses API
response = client.responses.create(
    model="google/gemini-3.1-pro-preview", 
    input="What is the meaning of life?"
)

print(response.output_text)

TypeScript

import OpenAI from "openai";

const openai = new OpenAI({
  baseURL: "https://api.agipower.ai/v1", 
  apiKey: "<your AGIPower_API_KEY>", 
});

async function main() {
  const response = await openai.responses.create({
    model: "google/gemini-3.1-pro-preview", 
    input: "What is the meaning of life?", 
  });

  console.log(response.output_text);
}

main();

cURL

curl https://api.agipower.ai/v1/responses \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer $AGIPower_API_KEY" \
  -d '{
    "model": "google/gemini-3.1-pro-preview",
    "input": "What is the meaning of life?"
  }'

Protocol 3: Anthropic Messages

AGIPower fully supports the Anthropic Messages protocol and integrates seamlessly with tools like Claude Code and Cursor.Note that the Anthropic protocol uses base_url https://agipower.ai, which differs from the OpenAI protocol.

Models compatible with the Anthropic protocol are being adapted in batches. You can view currently supported models by filtering for Anthropic API Compatible on the official model list: anthropic-support

You can also check on a model’s detail page: anthropic-support

Python

from anthropic import Anthropic

# 1. Initialize the Anthropic client
client = Anthropic(
    # 2. Point the base URL to the AGIPower endpoint
    base_url="https://agipower.ai", 
    # 3. Replace with the API key from your AGIPower console
    api_key="<your AGIPower_API_KEY>", 
)

# 4. Create a Messages request
message = client.messages.create(
    model="google/gemini-3.1-pro-preview", 
    max_tokens=1024,
    messages=[
        {
            "role": "user",
            "content": "What is the meaning of life?"
        }
    ]
)

print(message.content[0].text)

TypeScript

import Anthropic from "@anthropic-ai/sdk";

// 1. Initialize the Anthropic client
const client = new Anthropic({
  // 2. Point the base URL to the AGIPower endpoint
  baseURL: "https://agipower.ai", 
  // 3. Replace with the API key from your AGIPower console
  apiKey: "<your AGIPower_API_KEY>", 
});

async function main() {
  // 4. Create a Messages request
  const message = await client.messages.create({
    model: "google/gemini-3.1-pro-preview", 
    max_tokens: 1024,
    messages: [
      {
        role: "user",
        content: "What is the meaning of life?",
      },
    ],
  });

  console.log(message.content[0].text);
}

main();

cURL

curl https://api.agipower.ai/v1/messages \
  -H "content-type: application/json" \
  -H "x-api-key: $AGIPower_API_KEY" \
  -H "anthropic-version: 2023-06-01" \
  -d '{
    "model": "google/gemini-3.1-pro-preview",
    "max_tokens": 1024,
    "messages": [
      {
        "role": "user",
        "content": "What is the meaning of life?"
      }
    ]
  }'

Protocol 4: Google Gemini

AGIPower supports the Google Gemini (Vertex AI) protocol, allowing you to call models directly with the Google GenAI SDK.The base URL is https://api.agipower.ai. You need to set vertexai=True and configure custom http_options.

Python

from google import genai
from google.genai import types

# 1. Initialize the Google GenAI client
client = genai.Client(
    # 2. Replace with the API key from your AGIPower console
    api_key="<your AGIPower_API_KEY>", 
    vertexai=True, 
    # 3. Point the base URL to the AGIPower endpoint
    http_options=types.HttpOptions(
        api_version='v1',
        base_url='https://api.agipower.ai'
    )
)

# 4. Create a generateContent request
response = client.models.generate_content(
    model="google/gemini-3.1-pro-preview", 
    contents="What is the meaning of life?"
)

print(response.text)

TypeScript

import { GoogleGenAI } from "@google/genai";

// 1. Initialize the Google GenAI client
const ai = new GoogleGenAI({
  // 2. Replace with the API key from your AGIPower console
  apiKey: "<your AGIPower_API_KEY>", 
  vertexai: true, 
  // 3. Point the base URL to the AGIPower endpoint
  httpOptions: {
    apiVersion: "v1",
    baseUrl: "https://api.agipower.ai", 
  },
});

async function main() {
  // 4. Create a generateContent request
  const response = await ai.models.generateContent({
    model: "google/gemini-3.1-pro-preview", 
    contents: "What is the meaning of life?",
  });

  console.log(response.text);
}

main();

Platform API

AGIPower provides a set of Platform Management APIs that let you programmatically query account usage, balance, subscription status, and more — making it easy to integrate with your monitoring and management systems.

Platform APIs require a Management API Key for authentication (some endpoints also accept a regular API key). Go to AGIPower Console > Management to create a Management API Key.

Query Generation Details

Retrieve detailed information about a specific API call, including token usage, latency, and billing breakdown.

GET https://api.agipower.ai/v1/management/generation?id=<generation_id>

Each API call returns a generation_id. Use that ID to query the full details of the call:

curl https://api.agipower.ai/v1/management/generation?id=gen_01abc123def456 \
  -H "Authorization: Bearer $AGIPower_API_KEY"

Example response (partial):

{
  "api": "chat.completions",
  "generationId": "gen_01abc123def456",
  "model": "openai/gpt-4o",
  "generationTime": 3200,
  "latency": 500,
  "nativeTokens": {
    "completion_tokens": 128,
    "prompt_tokens": 32,
    "total_tokens": 160
  },
  "usage": 0.0052
}

Billing information (such as usage and ratingResponses) becomes available 3–5 minutes after the request completes. Token usage metrics are returned synchronously with the request.

For detailed parameter descriptions, see the Get Generation API docs.

Query PAYG Balance

Retrieve the current Pay As You Go balance for your account.

GET https://api.agipower.ai/v1/management/payg/balance

curl https://api.agipower.ai/v1/management/payg/balance \
  -H "Authorization: Bearer $AGIPower_MANAGEMENT_API_KEY"

Example response:

{
  "success": true,
  "data": {
    "currency": "usd",
    "total_credits": 482.74,
    "top_up_credits": 35.00,
    "bonus_credits": 447.74
  }
}

For detailed parameter descriptions, see the Get PAYG Balance API docs.

Query Subscription Details

Retrieve your account’s subscription information, including plan tier, Flow rate, and quota usage.

GET https://api.agipower.ai/v1/management/subscription/detail

curl https://api.agipower.ai/v1/management/subscription/detail \
  -H "Authorization: Bearer $AGIPower_MANAGEMENT_API_KEY"

Example response (partial):

{
  "success": true,
  "data": {
    "plan": {
      "tier": "ultra",
      "amount_usd": 200,
      "interval": "month"
    },
    "account_status": "healthy",
    "quota_5_hour": {
      "max_flows": 800,
      "used_flows": 57.2,
      "remaining_flows": 742.8,
      "usage_percentage": 0.0715
    },
    "quota_7_day": {
      "max_flows": 6182,
      "used_flows": 416.11,
      "remaining_flows": 5765.89
    }
  }
}

For detailed parameter descriptions, see the Get Subscription Detail API docs.

Query Flow Rate

Retrieve the current Flow rate information for your account.

GET https://api.agipower.ai/v1/management/flow_rate

curl https://api.agipower.ai/v1/management/flow_rate \
  -H "Authorization: Bearer $AGIPower_MANAGEMENT_API_KEY"

Example response:

{
  "success": true,
  "data": {
    "currency": "usd",
    "base_usd_per_flow": 0.03283,
    "effective_usd_per_flow": 0.03283
  }
}

For detailed parameter descriptions, see the Get Flow Rate API docs.

Advanced Usage

For more details on advanced usage, refer to the Advanced Usage section.

If you encounter any issues or have suggestions and feedback, feel free to reach out:

Official Website: https://agipower.ai
Technical Support: support@agipower.ai
Business Inquiries: support@agipower.ai
Twitter: @AGIPowerAI
Discord Community: http://discord.gg/vHZZzj84Bm

For more contact options and details, visit our Contact Us page.

​Quick Start

​Choose a Usage Plan

​Pay As You Go

​Builder Plan (Subscription)

​Plan Comparison

​Supported API Protocols

​Protocol 1: OpenAI Chat Completions

​Protocol 2: OpenAI Responses

​Protocol 3: Anthropic Messages

​Protocol 4: Google Gemini

​Platform API

​Query Generation Details

​Query PAYG Balance

​Query Subscription Details

​Query Flow Rate

​Advanced Usage

Quick Start

Choose a Usage Plan

Pay As You Go

Builder Plan (Subscription)

Plan Comparison

Supported API Protocols

Protocol 1: OpenAI Chat Completions

Protocol 2: OpenAI Responses

Protocol 3: Anthropic Messages

Protocol 4: Google Gemini

Platform API

Query Generation Details

Query PAYG Balance

Query Subscription Details

Query Flow Rate

Advanced Usage