Tether's AI research division has released QVAC BotPsy, an open-weight model with 800 million parameters designed to run locally on personal devices such as smartphones. Built on Alibaba's open-source Qwen3.5 architecture, the model converts user requests into tool calls to handle everyday tasks including schedules and messages. In benchmark testing across six agentic evaluations, BotPsy outperformed all tested models with no more than 1.2 billion parameters. In four core tests—BFCL v4, ACEBench, τ-bench, and ToolSandbox—it surpassed Qwen3.5-2B-Think, a model with roughly 2.5 times its parameter count, achieving tool-calling scores approximately twice those of its base model. The model weights are now available on Hugging Face.