We have launch a live-conversation product china, its a combination of- context-conversation-content producing model under 3 prepare phase：preparation / live/ after. We design the apps taking advantage of the thought of abstract painter Mondrian and wrack the feed, only allowing at most 7 people participate the show which must set a start time. The live-interface just like a painting allowing synchronize communication possible by making speakers word bubble up as bigger square than other speakers. So far we have driven over 3000 topic conversation，each will last 50min average and produced almost 8000 words back-and-forth casual content. Those content are circulated to other reading-recommendation machine and has won over 0.8 billion pageview, which intends the model of content producing will last and scale