# token-hub > LLM API gateway for OpenAI Chat Completions and Anthropic Messages through one TokenHub key. Pay-as-you-go from $5. No subscription. ## What - OpenAI-compatible HTTPS relay. Point an OpenAI SDK at the token-hub base URL and change the API key. - Anthropic Messages-compatible route for clients such as Claude Code. - One TokenHub key and one shared balance for enabled model channels. - Current runnable examples use the enabled public model route. ## Current routes - Base URL: https://llm.sandboxclaw.com/v1 - Chat: POST /chat/completions with `model: "moonshot-v1-8k"` - Anthropic Messages: POST /messages with `model: "moonshot-v1-8k"` - Streaming on Chat Completions is supported when the upstream channel supports it. - Responses API for Codex is reserved but not yet end-to-end usable on the live channel. ## Auth - Header: `Authorization: Bearer $TOKENHUB_KEY` - Use the full key generated in the console. Do not assume a fixed key prefix. - Keep the key server-side; do not expose it in browser code. ## Pricing - Pay-as-you-go topup starting at $5. - Public checkout is USD-only through Stripe. - Credits are valid for 12 months. - No subscription. Unused credits are refundable within 7 days. ## CLI support - Claude Code: set `ANTHROPIC_BASE_URL=https://llm.sandboxclaw.com` and `ANTHROPIC_API_KEY=$TOKENHUB_KEY`; the client reaches `/v1/messages`. - OpenCode: configure an OpenAI-compatible provider with `baseURL: "https://llm.sandboxclaw.com/v1"` and model `moonshot-v1-8k`. - DeepSeek-TUI: set `DEEPSEEK_API_KEY=$TOKENHUB_KEY`, `DEEPSEEK_BASE_URL=https://llm.sandboxclaw.com/v1`, and `DEEPSEEK_MODEL=moonshot-v1-8k`. - Codex: configure TokenHub as a Responses provider; model availability follows the API key's enabled model list. ## Docs - Quickstart: https://tokenhub.sandboxclaw.com/docs/quickstart - FAQ: https://tokenhub.sandboxclaw.com/docs/faq - Troubleshooting: https://tokenhub.sandboxclaw.com/docs/troubleshooting ## Contact - Website: https://tokenhub.sandboxclaw.com - Support: support@sandboxclaw.com