Technical Details

Deep dive into MyTokenGate’s architecture and implementation.

System Architecture

Gateway Service

MyTokenGate uses a high-performance gateway service built on:

• NestJS - Scalable Node.js framework
• PostgreSQL - Reliable data storage
• Redis - Fast caching and rate limiting

Request Flow

Client → Gateway → Authentication → Rate Limiting → Model Router → Provider API

Protocol Support

OpenAI Protocol

Full compatibility with OpenAI’s chat completions API:

• /v1/chat/completions
• /v1/completions
• /v1/embeddings
• /v1/models

Anthropic Protocol

Native support for Claude’s Messages API:

• /v1/messages
• Automatic prompt conversion
• Tool use and function calling

Rate Limiting

Token Bucket Algorithm

• Configurable requests per minute
• Token-based rate limiting
• Graceful degradation

Headers

X-RateLimit-Limit: 100
X-RateLimit-Remaining: 95
X-RateLimit-Reset: 1682524800

Billing System

Real-time Metering

• Token counting per request
• Model-specific pricing
• Usage aggregation

Invoice Generation

• Daily usage reports
• Monthly invoices
• Cost allocation by project

High Availability

Load Balancing

• Multiple provider endpoints
• Automatic failover
• Health checking

Data Redundancy

• Database replication
• Redis clustering
• Geographic distribution