# TokenReply > TokenReply is an API made for roleplay. You'll find the models roleplay players run, each with a Roleplay score and a Value score. Some are free, and repeated history is billed at the cache price on models that have one. JanitorAI, SillyTavern, RisuAI and other chat apps connect with one API key. Claude Code, Codex, Gemini CLI and the OpenAI SDKs work too: change the base URL and the key. - API origin: https://api.tokenreply.com - Base URL: https://api.tokenreply.com/v1 - Chat endpoint (the Proxy URL for JanitorAI and RisuAI): https://api.tokenreply.com/v1/chat/completions - Auth: `Authorization: Bearer sk-...` (create keys in the console) - Models and prices: https://www.tokenreply.com/models - Plans: Default is free and covers the free models. Paid models, including up to 40 requests a day to the models of the week (10 more on check-in days), are charged to the wallet. Plus is $5 a month or $50 a year and has no daily cap. Details: https://www.tokenreply.com/docs/plans ## Roleplay - [Roleplay API](https://www.tokenreply.com/roleplay-api): scored models and free options. Sections: "Roleplay scores", "Context, cache and presets" (why long chats cost more, how cache lowers it, and how TokenReply brings sampler values into the range each model accepts), "Plans", "Set up your app" (the setting to paste into JanitorAI, SillyTavern, RisuAI and other apps), "Troubleshooting" and "Frequently asked questions" - [JanitorAI API setup](https://www.tokenreply.com/janitorai-api): proxy setup with TokenReply. Sections: "Step 1: Create an account and an API key", "Step 2: Add TokenReply as a proxy in JanitorAI" (Proxy URL `https://api.tokenreply.com/v1/chat/completions`), "Step 3: Choose a model", "Recommended JanitorAI settings", "Keep streaming on", "What a message costs", "Troubleshooting" and "Frequently asked questions" - [SillyTavern API setup](https://www.tokenreply.com/sillytavern-api): Chat Completion with the Custom (OpenAI-compatible) source and base URL `https://api.tokenreply.com/v1`. Sections: "Step 1: Create an account and an API key", "Step 2: Connect SillyTavern", "Step 3: Context size, response length and streaming", "Step 4: Choose a model", "What a message costs", "Troubleshooting" and "Frequently asked questions" - [Roleplay costs](https://www.tokenreply.com/rankings/roleplay-cost): what 100 messages cost on every model, in a typical chat and a long one - [Free AI API](https://www.tokenreply.com/guides/free-ai-api): the models that are free right now ## Getting started - [Docs home](https://www.tokenreply.com/docs): accounts, keys, billing and tools - [Registration and sign-in](https://www.tokenreply.com/docs/registration) - [API key management](https://www.tokenreply.com/docs/create-key): create keys and set limits - [Using API keys](https://www.tokenreply.com/docs/use-api-key): the API URL and examples in Python, Node.js and cURL - [Plans and tiers](https://www.tokenreply.com/docs/plans): Default, Trial and Plus - [Billing and balance](https://www.tokenreply.com/docs/recharge): what gets charged first and how to top up - [Daily check-in](https://www.tokenreply.com/docs/check-in): free usage allowance every day ## API reference - [Base URL and authentication](https://www.tokenreply.com/docs/base-url) - [Compatibility](https://www.tokenreply.com/docs/compatibility): supported endpoints (`/v1/chat/completions`, `/v1/responses`, `/v1/messages`, `/v1/embeddings`, `/v1/rerank`, audio, images) and the apps you can connect - [Models and rules](https://www.tokenreply.com/docs/models) - [Code examples](https://www.tokenreply.com/docs/examples) - [FAQ and troubleshooting](https://www.tokenreply.com/docs/faq): errors by HTTP status code ## Coding tools - [Claude Code setup](https://www.tokenreply.com/docs/claude-code) - [Codex setup](https://www.tokenreply.com/docs/codex) - [Gemini CLI setup](https://www.tokenreply.com/docs/gemini-cli) - [CC-Switch](https://www.tokenreply.com/docs/cc-switch) - [Install Node.js](https://www.tokenreply.com/docs/node-setup) - [Third-party apps](https://www.tokenreply.com/docs/third-party-apps) ## Guides, comparisons and references - [Guides](https://www.tokenreply.com/guides): setup guides for SillyTavern, JanitorAI, DeepSeek, Grok, image generation and free API access - [Platforms](https://www.tokenreply.com/platforms): setup for each roleplay, chat and coding app - [TokenReply vs OpenRouter](https://www.tokenreply.com/compare/openrouter) - [OpenRouter alternatives](https://www.tokenreply.com/compare/openrouter-alternatives) - [Cheapest Claude API options](https://www.tokenreply.com/compare/cheapest-claude-api) - [LLM price calculator](https://www.tokenreply.com/tools/llm-price-calculator) - [Integrations](https://www.tokenreply.com/integrations): LangChain, LiteLLM, Cline / Roo Code, Dify, n8n, Anthropic SDK - [API error reference](https://www.tokenreply.com/errors): causes and fixes by HTTP status and error code - [Models by task](https://www.tokenreply.com/models/by-task) and [price rankings](https://www.tokenreply.com/rankings) - Every model has a page at `https://www.tokenreply.com/models/`; the full list is in https://www.tokenreply.com/sitemap-growth.xml ## Policies - [Terms of Service](https://www.tokenreply.com/terms-of-service) - [Privacy Policy](https://www.tokenreply.com/privacy-policy) - [Usage Policy](https://www.tokenreply.com/usage-policy) - [Refund Policy](https://www.tokenreply.com/refund-policy) - [Supported countries and territories](https://www.tokenreply.com/supported-countries-and-territories)