> ## Documentation Index
> Fetch the complete documentation index at: https://docs.agipower.ai/llms.txt
> Use this file to discover all available pages before exploring further.
# Quickstart
# Quick Start
Welcome to **[AGIPower](https://agipower.ai/about/intro.html)**! 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:
1. **Sign in to AGIPower**: Visit the **[AGIPower login page](https://agipower.ai/signin)** and sign in using any of the following:
* Email
* GitHub account
* Google account
2. **Choose a usage plan**: Based on your use case, select [Pay As You Go](#choose-a-usage-plan) or [Builder Plan](#choose-a-usage-plan).
3. **Get an API key**: After signing in, create an API key from the corresponding management page.
4. **Make your first request**: Pick the [API protocol](#supported-api-protocols) 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](https://agipower.ai/platform/pay-as-you-go)** to top up your balance and create an API key.
For complete instructions, see the [Pay As You Go Guide](/en/guide/pay-as-you-go).
### 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](https://agipower.ai/pricing/subscription)** to view plan details and subscribe.
For complete instructions, see the [Subscription Plans Guide](/en/guide/subscription).
### 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](https://agipower.ai/models):
Or on a [model's detail page](https://agipower.ai/google/gemini-3.1-pro-preview):
***
## 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 title="Python" theme={null}
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", # [!code highlight]
# 3. Replace with the API key from your AGIPower console
api_key="", # [!code highlight]
)
# 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", # [!code highlight]
messages=[
{
"role": "user",
"content": "What is the meaning of life?"
}
]
)
print(completion.choices[0].message.content)
```
```ts title="TypeScript" theme={null}
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", // [!code highlight]
// 3. Replace with the API key from your AGIPower console
apiKey: "", // [!code highlight]
});
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", // [!code highlight]
messages: [
{
role: "user",
content: "What is the meaning of life?",
},
],
});
console.log(completion.choices[0].message.content);
}
main();
```
```bash title="cURL" theme={null}
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 title="Python" theme={null}
from openai import OpenAI
client = OpenAI(
base_url="https://api.agipower.ai/v1", # [!code highlight]
api_key="", # [!code highlight]
)
# Create a request using the Responses API
response = client.responses.create(
model="google/gemini-3.1-pro-preview", # [!code highlight]
input="What is the meaning of life?" # [!code highlight]
)
print(response.output_text)
```
```ts title="TypeScript" theme={null}
import OpenAI from "openai";
const openai = new OpenAI({
baseURL: "https://api.agipower.ai/v1", // [!code highlight]
apiKey: "", // [!code highlight]
});
async function main() {
const response = await openai.responses.create({
model: "google/gemini-3.1-pro-preview", // [!code highlight]
input: "What is the meaning of life?", // [!code highlight]
});
console.log(response.output_text);
}
main();
```
```bash title="cURL" theme={null}
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](https://agipower.ai/models):
You can also check on a [model's detail page](https://agipower.ai/anthropic/claude-haiku-4.5):
```python title="Python" theme={null}
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", # [!code highlight]
# 3. Replace with the API key from your AGIPower console
api_key="", # [!code highlight]
)
# 4. Create a Messages request
message = client.messages.create(
model="google/gemini-3.1-pro-preview", # [!code highlight]
max_tokens=1024,
messages=[
{
"role": "user",
"content": "What is the meaning of life?"
}
]
)
print(message.content[0].text)
```
```ts title="TypeScript" theme={null}
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", // [!code highlight]
// 3. Replace with the API key from your AGIPower console
apiKey: "", // [!code highlight]
});
async function main() {
// 4. Create a Messages request
const message = await client.messages.create({
model: "google/gemini-3.1-pro-preview", // [!code highlight]
max_tokens: 1024,
messages: [
{
role: "user",
content: "What is the meaning of life?",
},
],
});
console.log(message.content[0].text);
}
main();
```
```bash title="cURL" theme={null}
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 title="Python" theme={null}
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="", # [!code highlight]
vertexai=True, # [!code highlight]
# 3. Point the base URL to the AGIPower endpoint
http_options=types.HttpOptions(
api_version='v1',
base_url='https://api.agipower.ai' # [!code highlight]
)
)
# 4. Create a generateContent request
response = client.models.generate_content(
model="google/gemini-3.1-pro-preview", # [!code highlight]
contents="What is the meaning of life?"
)
print(response.text)
```
```ts title="TypeScript" theme={null}
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: "", // [!code highlight]
vertexai: true, // [!code highlight]
// 3. Point the base URL to the AGIPower endpoint
httpOptions: {
apiVersion: "v1",
baseUrl: "https://api.agipower.ai", // [!code highlight]
},
});
async function main() {
// 4. Create a generateContent request
const response = await ai.models.generateContent({
model: "google/gemini-3.1-pro-preview", // [!code highlight]
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](https://agipower.ai/platform/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=
```
Each API call returns a `generation_id`. Use that ID to query the full details of the call:
```bash theme={null}
curl https://api.agipower.ai/v1/management/generation?id=gen_01abc123def456 \
-H "Authorization: Bearer $AGIPower_API_KEY"
```
Example response (partial):
```json theme={null}
{
"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](/en/api/platform/get-generation).
***
### Query PAYG Balance
Retrieve the current Pay As You Go balance for your account.
```
GET https://api.agipower.ai/v1/management/payg/balance
```
```bash theme={null}
curl https://api.agipower.ai/v1/management/payg/balance \
-H "Authorization: Bearer $AGIPower_MANAGEMENT_API_KEY"
```
Example response:
```json theme={null}
{
"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](/en/api/platform/payg-balance).
***
### 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
```
```bash theme={null}
curl https://api.agipower.ai/v1/management/subscription/detail \
-H "Authorization: Bearer $AGIPower_MANAGEMENT_API_KEY"
```
Example response (partial):
```json theme={null}
{
"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](/en/api/platform/subscription-detail).
***
### Query Flow Rate
Retrieve the current Flow rate information for your account.
```
GET https://api.agipower.ai/v1/management/flow_rate
```
```bash theme={null}
curl https://api.agipower.ai/v1/management/flow_rate \
-H "Authorization: Bearer $AGIPower_MANAGEMENT_API_KEY"
```
Example response:
```json theme={null}
{
"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](/en/api/platform/flow-rate).
***
## 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](https://agipower.ai)
* **Technical Support**: [support@agipower.ai](mailto:support@agipower.ai)
* **Business Inquiries**: [support@agipower.ai](mailto:support@agipower.ai)
* **Twitter**: [@AGIPowerAI](https://twitter.com/AGIPowerAI)
* **Discord Community**: [http://discord.gg/vHZZzj84Bm](http://discord.gg/vHZZzj84Bm)
For more contact options and details, visit our [Contact Us page](/en/help/contact).