Jatevo is an OpenAI-compatible inference gateway that aggregates multiple model providers and GPU resource pools, offering unified API access to various models. It supports two inference paths: Serverless (managed infrastructure) and Decentralized (community-contributed GPUs), with the latter using $JTVO tokens for payment or earning. Developers simply point their client's base URL to Jatevo and swap the API key to use standard OpenAI SDKs with different models (e.g., GPT, Claude, Gemini, Qwen). The platform also supports private deployments, allowing enterprises to connect the control plane to their own GPU pools. Its core mechanism routes requests to the optimal provider based on model ID, handling quotas, billing, and fault tolerance.