Call the /v1/chat/completions endpoint with your key in the Authorization Bearer header, the model ID in the model field, and the conversation so far in messages. Set stream to true when you want tokens streamed back. Multi-round chats reuse the same request shape with the full message history included.

Context: TaskingAI's OpenAI-compatible API template documents the stateless chat path. Chat completions go to /v1/chat/completions with an Authorization Bearer [redacted], taking model, stream, and messages fields; the TaskingAI model ID goes in the model field. The template shows both single-round and multi-round conversations using the same request shape.