BotPsy: On-Device Personal Assistant with a Sub-Billion-Parameter Agentic Language Model
Which summary reads better? Pick one — models revealed after.Both summaries are AI-generated.
An 800M-parameter agentic language model, BotPsy, achieves state-of-the-art tool-use performance for its size, enabling on-device deployment of personal assistants for tasks like web search and device control, reducing reliance on costly cloud LLMs and improving data privacy for users. This shift allows for cheaper and more private processing of routine workloads. It breaks the need for frontier-level intelligence for simple tasks.
BotPsy is a 0.8B-parameter Qwen3.5-based agent model for on-device function calling, with released safetensors and GGUF builds including bf16, Q8, and Q4. For production teams, the practical implication is that low/medium-complexity assistant workflows like search, calendar, email, files, commerce, and device control can move off frontier APIs to private, cheaper local inference, but multi-step information retrieval and high-stakes tasks still need routing to stronger models or guardrailed fallback.