Tether AI Research releases BotPsy 0.8B on-device agentic model
Start here
Tether AI Research has released BotPsy, a 0.8B-parameter agentic model built on Qwen3.5-0.8B and intended for on-device personal-assistant tool use. According to the company, it roughly doubles its base model's agentic scores, with BFCL v4 rising from 23.4 to 44.0 and τ-bench from 23.2 to 51.5, and it beats every model up to 1.2B parameters tested across six tool-use benchmarks. Tether also reports unsafe responses falling to 21% from 36% and about 3× fewer reasoning tokens on GPQA Diamond. The weights are to be released under the Apache 2.0 license for research and educational use, in bf16, Q8 and Q4 precisions as Hugging Face safetensors and GGUF files for the QVAC SDK and llama.cpp. Tether says the small parameter count suits it best to low- to medium-difficulty tasks, and that it was primarily trained and tested in English, the recommended language.
Generated from reports · updated 22 h ago
Report timeline
Follow the reports to see the event from different sides.
- Hugging Face BlogTether AI Research Releases BotPsy: A 0.8B On-Device Agentic Model Built on Qwen3.5-0.8B
Tether AI Research released BotPsy, a 0.8B-parameter agentic model built on Qwen3.5-0.8B for on-device personal-assistant tool use. It roughly doubles its base model's agentic scores, with BFCL v4 rising from 23.4 to 44.0 and τ-bench from 23.2 to 51.5, and beats every model up to 1.2B parameters tested across six tool-use benchmarks. It also cuts unsafe responses to 21% from 36% and uses about 3× fewer reasoning tokens on GPQA Diamond.
Heat for this event
Not enough continuous observations yet to draw a trend.