TokensMind

TokensMind is an LLM gateway that exposes one OpenAI-compatible endpoint and routes requests to 100+ models from providers including OpenAI, Anthropic, Google, DeepSeek, Qwen, Kimi, MiniMax, and Zhipu. A single API key and base URL cover chat, image, video, speech-to-text, embeddings, and reranking.

It adds automatic model routing, cost tracking, spending limits, and observability dashboards for cost, latency, error rates, and cache hits, plus role-based access control, audit logs, and guardrails. Billing is pay-as-you-go with transparent per-request pricing.

It integrates with tools like Cursor and Claude Code via the OpenAI-compatible API and MCP, so developers can point existing clients at one endpoint instead of wiring up each provider.