FrontierThe story, in brief
Tencent's Gander aims to keep talking while it works in the background
Tencent's Gander splits the brain: one part keeps you talking, another handles the work. Interruptions drop to 8%—but accuracy still lags.
The KeyNews take
Why it matters
A novel architecture for multimodal agents that separates conversational continuity from task execution. Demonstrates a real capability tradeoff: smoother UX vs. task reliability. Relevant to how frontier labs are rethinking agent design.
The key facts
5 to knowTencent's Gander uses 'cerebellum' for speech/conversation and swappable 'brain' for background tasks
Handles speech, images, and text multimodally
8% interrupt rate in benchmarks, better than rivals
Task accuracy trails competitors
Users can interrupt or redirect mid-conversation
Go to the source
The Decoderthe-decoder.com
Publisher excerpt: Tencent's Gander processes speech, images, and text while handling tasks in the background. A "cerebellum" keeps the conversation going, while a swappable "brain" searches files, writes code, or tackles other complex work. Users can interrupt or change the task mid-conversation. In benchmarks,…