👋 Welcome to TokenReply
TokenReply is an AI API gateway that aggregates 40+ upstream providers (OpenAI, Claude, Gemini, Azure, AWS Bedrock, etc.) behind a unified OpenAI-compatible interface. For most clients, getting started is a matter of changing the base URL, API key, and model name.
🌟 Core Capabilities
- Multi-provider aggregation: access OpenAI, Claude, Gemini, Llama, and more through a single API, no per-provider integration needed
- Billing & limit management: in-plan usage, wallet balance, API key-level USD spending limits, per-model pricing, and Plan-based access to different routes
- Rate limiting: multi-layer rate limiting across the account, API key, and route; Plus gets a higher rate limit
- OpenAI compatibility: supports Chat Completions, Responses, Embeddings, Images, Audio, and more
- Streaming: full SSE streaming support for low-latency real-time responses
🚀 Quick Start
Three steps to get started:
- Create an account: via email or OAuth
- Create an API key: create a key on the API Key Management page and choose an available Channel
- Use your API key: set the API URL to https://api.tokenreply.com/v1, then provide the API key and model name
💡 Tip
If your client already speaks the OpenAI format, no request body changes are needed. Focus on getting the base URL, Authorization header, and model name right.
A new account gets a one-month Trial allowance. To keep calling, claim free allowance with daily check-in, top up your wallet, or upgrade to Plus.
📚 Documentation
User Guide
- Registration & Login: account creation, email verification, OAuth
- Using API Key: start with API URL https://api.tokenreply.com/v1, then integrate with Python, Node.js, cURL, and more
- API Key Management: API key creation, Channel selection, restrictions, and rotation
- Plans & Tiers: Default / Trial / Plus and how to upgrade
- Billing & Balance: in-plan usage, wallet balance, top-up, billing records
- Daily Check-in: claim free in-plan usage daily
- Referral Rewards: earn allowance and free Plus by inviting friends
API Reference
- Base URL & Auth: SDK URL, raw endpoints, headers, and response conventions
- Compatibility: supported endpoints, tested clients
- Models & Rules: model availability, naming, pricing
AI Coding Tools
- Install Node.js: required runtime for AI coding tools
- Claude Code Setup: Anthropic's CLI coding tool
- CodeX Setup: OpenAI's GPT coding assistant
- Gemini CLI Setup: Google's AI coding assistant
Tools & Apps
- CC-Switch Config Tool: GUI tool for managing AI tool configs
- Third-Party Apps: CherryStudio, OpenCode, and other clients
Help
- Code Examples: streaming, raw HTTP, embeddings, and other variants
- FAQ & Troubleshooting: common errors and solutions
- Models: browse all available models and pricing