AK AI CORE API Documentation

OpenAI-compatible API for 140+ AI models. One endpoint, one wallet, instant access to GPT-4o, Claude 3.5, Gemini, and more. Pay-as-you-go with bKash/Nagad.

AK AI CORE API

Welcome to the AK AI CORE API documentation. Our API is OpenAI-compatible, letting you access 140+ AI models through a single endpoint with pay-as-you-go billing via bKash/Nagad.

AK AI CORE provides a unified gateway to the world's best AI models. Instead of managing multiple API keys from OpenAI, Anthropic, Google, and others, you get one endpoint, one wallet, and instant access to GPT-4o, Claude 3.5, Gemini, Llama, and more.

OpenAI-compatible API — drop-in replacement for existing integrations.

140+ models from OpenAI, Anthropic, Google, Meta, Alibaba, and more.

Pay-as-you-go billing with local Bangladeshi payment methods (bKash, Nagad, UddoktaPay).

No monthly subscriptions — top up credits and use any model instantly.

Single API key for all models. No need to manage multiple provider accounts.

Base URL

bash

https://akaicore.space/v1

All API requests should be made to this base URL. The API follows standard REST conventions and returns JSON responses.

Quick Example

bash

1curl -X POST https://akaicore.space/v1/chat/completions \
2  -H "Authorization: Bearer YOUR_API_KEY" \
3  -H "Content-Type: application/json" \
4  -d '{
5    "model": "openai/gpt-4o-mini",
6    "messages": [
7      {"role": "user", "content": "Hello from AK AI CORE"}
8    ]
9  }'

Authentication

AK AI CORE uses API key authentication via the Authorization header, fully compatible with the OpenAI SDK. Generate your API key from the dashboard and pass it in every request.

All API requests require an API key passed in the Authorization header as a Bearer token. You can generate API keys from your AK AI CORE dashboard under Settings > API Keys.

API Key Format

bash

Authorization: Bearer sk-or-xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx

cURL Example

bash

curl https://akaicore.space/v1/models \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json"

Python (OpenAI SDK)

python

1from openai import OpenAI
2
3client = OpenAI(
4    api_key="YOUR_API_KEY",  # Your AK AI CORE API key
5    base_url="https://akaicore.space/v1"
6)
7
8# Now use client exactly like OpenAI
9response = client.chat.completions.create(
10    model="openai/gpt-4o-mini",
11    messages=[{"role": "user", "content": "Hello"}]
12)

Node.js (OpenAI SDK)

javascript

1import OpenAI from "openai";
2
3const client = new OpenAI({
4  apiKey: "YOUR_API_KEY",  // Your AK AI CORE API key
5  baseUrl: "https://akaicore.space/v1"
6});
7
8// Now use client exactly like OpenAI
9const response = await client.chat.completions.create({
10  model: "openai/gpt-4o-mini",
11  messages: [{ role: "user", content: "Hello" }]
12});

Security Notes

API keys start with sk-or- prefix and should be treated as secrets.

Never expose your API key in client-side code or public repositories.

You can revoke and regenerate keys anytime from the dashboard.

Each API key is tied to your account credit balance.

Chat Completions

Send messages to any AI model and receive a completion. Fully compatible with the OpenAI Chat Completions API — just change the base_url.

POSThttps://akaicore.space/v1/chat/completions

cURL

bash

1curl -X POST https://akaicore.space/v1/chat/completions \
2  -H "Authorization: Bearer YOUR_API_KEY" \
3  -H "Content-Type: application/json" \
4  -d '{
5    "model": "openai/gpt-4o-mini",
6    "messages": [
7      {"role": "system", "content": "You are a helpful assistant."},
8      {"role": "user", "content": "What is the capital of Bangladesh?"}
9    ],
10    "temperature": 0.7,
11    "max_tokens": 512
12  }'

Python (OpenAI SDK)

python

1from openai import OpenAI
2
3client = OpenAI(
4    api_key="YOUR_API_KEY",
5    base_url="https://akaicore.space/v1"
6)
7
8response = client.chat.completions.create(
9    model="openai/gpt-4o-mini",
10    messages=[
11        {"role": "system", "content": "You are a helpful assistant."},
12        {"role": "user", "content": "What is the capital of Bangladesh?"}
13    ],
14    temperature=0.7,
15    max_tokens=512
16)
17
18print(response.choices[0].message.content)

Node.js (OpenAI SDK)

javascript

1import OpenAI from "openai";
2
3const client = new OpenAI({
4  apiKey: "YOUR_API_KEY",
5  baseUrl: "https://akaicore.space/v1"
6});
7
8const response = await client.chat.completions.create({
9  model: "openai/gpt-4o-mini",
10  messages: [
11    { role: "system", content: "You are a helpful assistant." },
12    { role: "user", content: "What is the capital of Bangladesh?" }
13  ],
14  temperature: 0.7,
15  max_tokens: 512
16});
17
18console.log(response.choices[0].message.content);

Request Body

json

1{
2  "model": "openai/gpt-4o-mini",
3  "messages": [
4    { "role": "system", "content": "You are a helpful assistant." },
5    { "role": "user", "content": "Hello!" }
6  ],
7  "temperature": 0.7,
8  "max_tokens": 512,
9  "stream": false,
10  "top_p": 1.0,
11  "frequency_penalty": 0.0,
12  "presence_penalty": 0.0
13}

Response Body

json

1{
2  "id": "chatcmpl-abc123",
3  "object": "chat.completion",
4  "created": 1718323200,
5  "model": "openai/gpt-4o-mini",
6  "choices": [
7    {
8      "index": 0,
9      "message": {
10        "role": "assistant",
11        "content": "The capital of Bangladesh is Dhaka."
12      },
13      "finish_reason": "stop"
14    }
15  ],
16  "usage": {
17    "prompt_tokens": 24,
18    "completion_tokens": 10,
19    "total_tokens": 34
20  }
21}

Streaming (SSE)

Set "stream": true to receive tokens in real-time via Server-Sent Events. Works identically to OpenAI's streaming API.

bash

1curl -X POST https://akaicore.space/v1/chat/completions \
2  -H "Authorization: Bearer YOUR_API_KEY" \
3  -H "Content-Type: application/json" \
4  -d '{
5    "model": "openai/gpt-4o-mini",
6    "messages": [{"role": "user", "content": "Hello"}],
7    "stream": true
8  }'

Request Parameters

model (required) — Model ID like openai/gpt-4o-mini, anthropic/claude-3-5-sonnet.

messages (required) — Array of {role, content} objects. Roles: system, user, assistant.

temperature (optional, default 0.7) — Sampling temperature between 0 and 2.

max_tokens (optional, default 2048) — Maximum tokens in the response.

stream (optional, default false) — Set to true for real-time SSE streaming.

top_p (optional, default 1.0) — Nucleus sampling parameter.

frequency_penalty (optional, default 0.0) — Reduce repetition.

presence_penalty (optional, default 0.0) — Encourage topic diversity.

Models

List all available models on AK AI CORE. Returns model IDs, pricing, and capabilities in OpenAI-compatible format.

GEThttps://akaicore.space/v1/models

Returns a list of all active models you can use. Model IDs follow the provider/model-name format (e.g., openai/gpt-4o-mini).

cURL

bash

curl https://akaicore.space/v1/models \
  -H "Authorization: Bearer YOUR_API_KEY"

Python

python

1from openai import OpenAI
2
3client = OpenAI(
4    api_key="YOUR_API_KEY",
5    base_url="https://akaicore.space/v1"
6)
7
8models = client.models.list()
9for model in models:
10    print(model.id)

Node.js

javascript

1import OpenAI from "openai";
2
3const client = new OpenAI({
4  apiKey: "YOUR_API_KEY",
5  baseUrl: "https://akaicore.space/v1"
6});
7
8const models = await client.models.list();
9console.log(models.data.map(m => m.id));

Response

json

1{
2  "object": "list",
3  "data": [
4    {
5      "id": "openai/gpt-4o-mini",
6      "object": "model",
7      "created": 1715644800,
8      "owned_by": "openai"
9    },
10    {
11      "id": "anthropic/claude-3-5-sonnet",
12      "object": "model",
13      "created": 1718323200,
14      "owned_by": "anthropic"
15    }
16  ]
17}

Wallet & Credits

AK AI CORE uses a simple pay-as-you-go credit system. Top up via bKash, Nagad, or UddoktaPay — then use any model instantly. No subscriptions, no hidden fees.

All API usage is billed through credits. $1 USD credit = $1 USD of usage. Credits are deducted in real-time based on actual token consumption and model pricing.

How Pricing Works

Each model has input and output prices per million tokens.

Cost = (prompt_tokens / 1M x input_price) + (completion_tokens / 1M x output_price).

Prices already include our platform margin — no surprise charges.

Micro-costs (like $0.0015) are tracked precisely and never rounded to $0.00.

Top Up Methods

bKash — Instant top-up via bKash Send Money.

Nagad — Top-up via Nagad payment.

UddoktaPay — BDT to USD conversion with competitive rates.

NOWPayments — Crypto payments for international users (BTC, ETH, USDT).

Checking Balance

View your current balance and transaction history in the AK AI CORE dashboard under Wallet.

json

1{
2  "balance": 12.45,
3  "transactions": [
4    {
5      "id": "tx_abc123",
6      "amount": -0.001520,
7      "type": "USAGE",
8      "description": "Chat: Claude 3.5 Sonnet",
9      "status": "success",
10      "createdAt": "2025-01-15T10:30:00Z"
11    },
12    {
13      "id": "tx_xyz789",
14      "amount": 10.00,
15      "type": "TOPUP",
16      "description": "Top-up via bKash",
17      "status": "success",
18      "createdAt": "2025-01-14T08:00:00Z"
19    }
20  ]
21}

Low Balance Alert

When your balance drops below $0.50, API requests will return 402 Payment Required. Top up anytime from the dashboard.

json

{
  "error": "Insufficient credits. Please top up your account."
}

Error Codes

All API errors follow a consistent JSON format with an error field. HTTP status codes indicate the type of failure.

json

{
  "error": "Human-readable error message"
}

HTTP Status Codes

Code	Meaning	Common Cause
200	OK	Request succeeded.
201	Created	Resource created (API key, account, etc).
400	Bad Request	Missing required fields, invalid payload, or validation failure.
401	Unauthorized	Invalid or expired API key, or missing session cookie.
402	Payment Required	Insufficient credits to process the request.
403	Forbidden	IP not in allowlist, admin-only endpoint, or user mismatch.
404	Not Found	Model, conversation, user, or resource does not exist.
409	Conflict	Email already registered or duplicate resource.
429	Too Many Requests	Rate limit exceeded. Implement backoff and retry.
500	Internal Server Error	Unexpected server failure. Contact support if persistent.
503	Service Unavailable	No active providers available to route the request.

Error Handling Example

typescript

1const response = await fetch('https://akaicore.space/v1/chat/completions', {
2  method: 'POST',
3  headers: {
4    'Content-Type': 'application/json',
5    'Authorization': 'Bearer YOUR_API_KEY'
6  },
7  body: JSON.stringify(payload),
8});
9
10if (!response.ok) {
11  const { error } = await response.json();
12
13  switch (response.status) {
14    case 401:
15      console.error('Unauthorized:', error);
16      break;
17    case 402:
18      console.log('Insufficient credits:', error);
19      break;
20    case 429:
21      console.log('Rate limited. Retrying after backoff...');
22      await new Promise(r => setTimeout(r, 2000));
23      break;
24    case 404:
25      console.log('Model not found:', error);
26      break;
27    case 500:
28      console.error('Server error:', error);
29      break;
30    default:
31      console.error('API error:', response.status, error);
32  }
33}