VaultGate

Unified AI Gateway
One endpoint for all your AI models

Website | Documentation | Dashboard | Chat

---

Overview

VaultGate is a production-ready API gateway that unifies 40+ AI providers into a single OpenAI-compatible endpoint. Access models from OpenAI, Anthropic, Google, Mistral, DeepSeek, Meta, Groq, and more through one consistent API.

---

Features

• OpenAI-Compatible API: Drop-in replacement for OpenAI's API format
• 100+ AI Models: Access text and image generation models from major providers
• Unified Endpoint: Single API for chat completions and image generation
• Real-time Streaming: Server-sent events for streaming responses
• API Key Management: Create keys with rate limits, token quotas, and model restrictions
• Interactive Playground: Test models directly in the browser
• Usage Analytics: Track requests, tokens, and costs per key
• Smart Failover: Automatic routing when primary providers are unavailable

---

Getting Started

Step 1: Create an Account

1. Visit endpoint.vaultgate.dev/auth
2. Click "Register" to create a new account
3. Enter your email, username, and password
4. After registration, you will be redirected to the dashboard

Step 2: Add Credits to Your Balance

1. Navigate to the "Balance" section in the sidebar
2. Add credits to your account using the available payment methods
3. Credits are used to pay for API requests based on token usage
4. You only pay for what you use, no subscriptions required

Step 3: Generate an API Key

1. Go to the "API Keys" page from the sidebar
2. Click "Create API Key"
3. Enter a name for your key (e.g., "My Application", "Development", "Production")
4. Configure optional settings:
   • Rate Limit: Maximum requests per minute (default: 60)
   • Token Quota: Maximum tokens allowed for this key
   • Model Restrictions: Limit which models this key can access
5. Click "Create" and copy your API key immediately
6. Your key starts with vg_ and will not be shown again after creation

Step 4: Browse Available Models

1. Visit the "Models" page to see all available AI models
2. Filter models by type:
   • Text: Chat and completion models (GPT-4, Claude, Gemini, etc.)
   • Image: Image generation models (DALL-E, Flux, Midjourney, etc.)
   • Hybrid: Models that support both text and image
3. Filter by provider: OpenAI, Anthropic, Google, Mistral, Meta, and more
4. Note the model ID for use in your API requests (e.g., gpt-4o-text, claude-3-5-sonnet-text)

Step 5: Test in the Playground

Before integrating into your application, test models in the browser:

1. Go to the "Playground" page
2. Select a model from the dropdown
3. Choose which API key to use for the request
4. Adjust settings:
   • Temperature: Controls randomness (0 = deterministic, 2 = creative)
   • Max Tokens: Limit response length
   • Streaming: Enable real-time response streaming
5. Type your message and send
6. View response time, token usage, and costs
7. Export conversations or copy generated code

Step 6: Make Your First API Request

Use the unified endpoint with your API key:

Chat Completion:

curl https://api.vaultgate.dev/v1/chat/completions \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer vg_your_api_key" \
  -d '{
    "model": "gpt-4o-text",
    "messages": [
      {"role": "system", "content": "You are a helpful assistant."},
      {"role": "user", "content": "Hello!"}
    ],
    "stream": true
  }'

Image Generation:

curl https://api.vaultgate.dev/v1/images/generations \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer vg_your_api_key" \
  -d '{
    "model": "dall-e-3-image",
    "prompt": "A futuristic city at sunset",
    "n": 1,
    "size": "1024x1024"
  }'

Step 7: Monitor Your Usage

1. Visit the "Logs" page to see all API requests
2. View details for each request:
   • Model used
   • Tokens consumed (prompt + completion)
   • Response time
   • Status (success/error)
3. Track spending and usage patterns over time
4. Set up alerts when approaching quota limits

---

Workflow Summary

1. Register Account     -->  endpoint.vaultgate.dev/auth
2. Add Credits          -->  Balance page
3. Create API Key       -->  API Keys page (save your vg_* key)
4. Browse Models        -->  Models page (find model IDs)
5. Test in Playground   -->  Playground page (optional but recommended)
6. Integrate API        -->  Use https://api.vaultgate.dev/v1
7. Monitor Usage        -->  Logs page

---

API Reference

Base URL

https://api.vaultgate.dev/v1

Authentication

All requests require a Bearer token in the Authorization header:

Authorization: Bearer vg_your_api_key

API keys are created from the dashboard and start with vg_. Each key can have its own rate limits and permissions.

Endpoints

Method	Endpoint	Description
GET	/v1/models	List all available models
POST	/v1/chat/completions	Generate chat completions (text)
POST	/v1/images/generations	Generate images from text prompts

Model Naming Convention

Model IDs include a type suffix to indicate their capabilities:

• -text for chat and completion models
• -image for image generation models
• -hybrid for models that support both text and image

Examples:

• gpt-4o-text - OpenAI GPT-4o for chat
• claude-3-5-sonnet-text - Anthropic Claude 3.5 Sonnet
• dall-e-3-image - OpenAI DALL-E 3 for images
• gemini-1.5-pro-hybrid - Google Gemini with multimodal support

Chat Completions Request

{
  "model": "gpt-4o-text",
  "messages": [
    {"role": "system", "content": "You are a helpful assistant."},
    {"role": "user", "content": "Your message here"}
  ],
  "temperature": 0.7,
  "max_tokens": 1000,
  "stream": true
}

Parameter	Type	Description
model	string	Required. Model ID from /v1/models
messages	array	Required. Array of message objects with role and content
temperature	number	Sampling temperature 0-2 (default: 1)
max_tokens	integer	Maximum tokens to generate
stream	boolean	Enable streaming responses (default: false)
top_p	number	Nucleus sampling parameter 0-1

Image Generation Request

{
  "model": "dall-e-3-image",
  "prompt": "A description of the image you want",
  "n": 1,
  "size": "1024x1024"
}

Parameter	Type	Description
model	string	Required. Image model ID
prompt	string	Required. Text description of the image
n	integer	Number of images to generate (default: 1)
size	string	Image dimensions (e.g., "1024x1024")

Error Codes

Status	Description
400	Bad Request - Invalid parameters
401	Unauthorized - Invalid or missing API key
402	Payment Required - Insufficient credits
404	Not Found - Model not available
429	Too Many Requests - Rate limit exceeded
503	Service Unavailable - Upstream provider error

Rate Limits

Rate limits are configured per API key. Default is 60 requests per minute. You can view and adjust limits in the API Keys page. When exceeded, you receive a 429 status code.

---

Supported Providers

Provider	Text Models	Image Models
OpenAI	GPT-4o, GPT-4, GPT-3.5	DALL-E 3
Anthropic	Claude 3.5, Claude 3	-
Google	Gemini Pro, Gemini Flash	Imagen
Mistral	Mistral Large, Codestral	-
Meta	Llama 3.1, Llama 3	-
DeepSeek	DeepSeek V3, DeepSeek Coder	-
Groq	Llama, Mixtral	-
Flux	-	Flux Pro, Flux Dev

View the complete list at endpoint.vaultgate.dev/models

---

SDK Integration

VaultGate works with any OpenAI-compatible SDK. Simply change the base URL:

Python (OpenAI SDK)

from openai import OpenAI

client = OpenAI(
    api_key="vg_your_api_key",
    base_url="https://api.vaultgate.dev/v1"
)

response = client.chat.completions.create(
    model="gpt-4o-text",
    messages=[{"role": "user", "content": "Hello!"}]
)

JavaScript/TypeScript

import OpenAI from 'openai';

const client = new OpenAI({
  apiKey: 'vg_your_api_key',
  baseURL: 'https://api.vaultgate.dev/v1'
});

const response = await client.chat.completions.create({
  model: 'gpt-4o-text',
  messages: [{ role: 'user', content: 'Hello!' }]
});

cURL

curl https://api.vaultgate.dev/v1/chat/completions \
  -H "Authorization: Bearer vg_your_api_key" \
  -H "Content-Type: application/json" \
  -d '{"model": "claude-3-5-sonnet-text", "messages": [{"role": "user", "content": "Hello!"}]}'

---

VaultGate Chat

VaultGate Chat is a full-featured chat interface for interacting with AI models:

• Access at chat.vaultgate.dev
• Connect using your VaultGate API key
• Chat with any available text model
• Generate images with image models
• Conversation history and management
• Model switching within conversations
• Markdown rendering and code syntax highlighting
• Export and share conversations

---

Screenshots

Portal

The landing page provides quick access to VaultGate Chat and the Endpoint dashboard.

Endpoint Landing

The main dashboard landing page with overview and quick actions.

Models

Browse and filter available AI models by provider and type.

API Keys

Create and manage API keys with rate limits and permissions.

Balance

Add credits and track your spending.

Profile

Manage your account settings and preferences.

Playground

Test any model directly in your browser with real-time streaming.

Logs

View all API requests, token usage, and response times.

Chat Interface

A full-featured chat interface for interacting with AI models.

Documentation

Comprehensive API documentation and integration guides.

---

Project Structure

This repository contains four main components:

VaultGate/
├── VaultGate.portal/     # Landing page (React + Vite)
├── VaultGate.chat/       # Chat interface (React + Vite)
├── VaultGate.endpoint/   # API gateway dashboard and backend
│   ├── gateway/          # Dashboard frontend (Next.js)
│   └── gateway-server/   # API server (Deno)
└── VaultGate.docs/       # Documentation site (React + Vite)

---

Technology Stack

Frontend

• React 19
• Next.js 16
• TypeScript
• Tailwind CSS
• Framer Motion

Backend

• Deno 2.0
• Turso (LibSQL)

Infrastructure

• Vercel (Frontend hosting)
• Deno Deploy / Railway (Backend)

---

Links

• Website: vaultgate.dev
• Documentation: docs.vaultgate.dev
• Dashboard: endpoint.vaultgate.dev
• Chat: chat.vaultgate.dev
• API Base URL: api.vaultgate.dev/v1

---

FAQ

How do I get started? Create an account at endpoint.vaultgate.dev, add credits, generate an API key, and start making requests.

Is VaultGate compatible with OpenAI SDKs? Yes. VaultGate is fully OpenAI-compatible. Change the base URL to https://api.vaultgate.dev/v1 and use your VaultGate API key.

What happens if a provider goes down? VaultGate includes smart failover. If a primary provider is unavailable, requests are automatically routed to backup providers when possible.

How is pricing calculated? You pay per token used. Different models have different costs. Check the Models page for pricing details.

Can I restrict which models an API key can access? Yes. When creating an API key, you can specify which models or model types the key is allowed to use.

Is there a free tier? New accounts receive starter credits to test the platform. After that, you pay only for what you use.

---

License

MIT License

---

_{Built with care by the VaultGate Team}