January 20, 2011 - BeijingSenseTimeLtd. today announced that its "SenseNova Fusion Big Model Interactive Edition" (SenseNova-5o) is now available to the public in real-time.audio-video dialogServices.Free use for a limited time.
According to the introduction, the model is the interactive version of the "daily new" fusion model of the Shangtang, supportingReal-time interaction, visual recognition, memory thinking, continuous dialog, and complex reasoningthat can help AI communicate more naturally and fluently with humans.
Shangtang also provides service optimization of the supporting Realtime API for "SenseNova-5o" to achieve integration with RTC networks. Officially, users in any environment canStable, real-time, smooth, latency-free audio and video dialog and communication.
1AI Attachment of the highlights information of ShangTech's "Day-to-Day Convergence Large Model Interactive Edition" is as follows:
- Supports ultra-long multimodal interaction memoryNot less than 5 minutes
- Continuously tracking and accumulating information on interactions with users, constantly refining and optimizing the understanding of user needs
- The current interaction latency is reduced to 2 secondsWithin, it claims to be "virtually indistinguishable from natural human communication."
- supportInterrupt at any time, keep the conversation going, and lead to new topics based on context
- Supports personalization of communication style and usage habits (persona, tone of voice, etc.) according to user preferences
- Support to help parents with their children's homework