Brave Leo Offers NEAR AI Inference When Users Select GLM 5.3 Flash

Brave Leo offers private inference through NEAR AI Cloud when users select GLM 5.3 Flash, tying encrypted cloud processing to a specific model choice in the browser assistant.

For those sessions, Leo encrypts prompts inside the browser before processing them in a verified hardware enclave. NEAR AI and the cloud host cannot read the prompt contents during inference. The privacy protections described apply to GLM 5.3 Flash sessions, not every Leo chat.