ByteDance has escalated China’s Lunar New Year tech wars by sending its AI assistant Doubao onto the main stage of the CCTV 2026 Spring Festival Gala. On the night of the gala Doubao will distribute cash red envelopes and hand out more than 100,000 technology prizes that are said to be integrated with Doubao’s large model via ByteDance’s Volcano Engine. The prize package spans 17 products from consumer electronics to mobility, including robots, a 3D printer, DJI drones, projectors and even short-term usage rights for two electric cars.
The move marks a strategic departure from rival campaigns that have focused on cash incentives or instant retail coupons. Tencent has leaned on social mechanics around Yuanbao with a reported billion-yuan push aimed at recreating the viral effect of earlier mobile-red-envelope moments, while Alibaba’s Qiwen has deployed large-scale consumption subsidies and integration with its commerce networks to drive immediate transactions. ByteDance’s approach is different: use a cultural touchpoint to seed hands-on experiences with a hardware ecosystem that speaks with Doubao’s AI.
ByteDance’s CEO Liang Rubo set a tone of aggressive ambition in the company’s January all-hands, naming “climbing to new heights” and singling out Doubao and the Dola assistant as short-term priorities. The Spring Gala appearance is effectively Doubao’s first large-scale hardware showing, and it comes after a sustained push into devices — from earbuds to last year’s high-profile Doubao smartphone — and persistent reports about forthcoming headsets and new audio models.
Technically, ByteDance says the hardware prizes will be fused with its large-model capabilities so that devices offer upgraded interaction: a home robot’s voice and conversational persona will be supplied by Doubao’s speech synthesis and language models, and the Audi E5 Sportback prize is promoted as offering an “Audi assistant” driven by the same model to enable near-human car dialogue. That pitch combines two current industry bets: richer multimodal models and the argument that on-device or edge-connected hardware is the best route to habitual AI use.
Shanghai-based digital economy scholars and industry observers view the campaign as ecosystem-building rather than a one-off marketing splash. Giving away hardware that runs on Doubao is a way to create persistent, contextual AI touchpoints in households — appliances, projectors or cars become ongoing portals for ByteDance’s assistant. The strategic logic is to convert a seasonal spike in users into longer-term habits and first-party data flows that can feed model improvement and vertical specialisation.
The gamble is long-term and costly. Hardware-driven adoption can produce richer interactions and stickier users, but it takes time and heavy investment to build supply chains, service ecosystems and after-sales support. Meanwhile, the Lunar New Year battlefield already shows different paths to a similar prize: Tencent’s social viral funnel, Alibaba’s commerce-first incentives, Baidu’s developer and cloud focus and Kuaishou’s content-to-local-services angle. Each path builds a different kind of “super‑entry point” for AI in everyday life.
There are also risks beyond cost and logistics. Bringing AI-enabled devices into millions of homes intensifies questions about data collection, user consent and regulatory scrutiny, especially where audio, vision and in-car telemetry are involved. Trust incidents or regulatory pushback could blunt the strategy’s gains; conversely, strong safety and privacy practices coupled with polished device experiences would put ByteDance in an advantaged position to monetise AI at the consumer edge.
For international observers, the episode illustrates how the battle for consumer AI in China is entering a materially different phase: competition is no longer only about cloud models and chat interfaces but about stitching AI into physical things people use daily. The Spring Gala is a staging ground, but the substantive contest will be fought over who can create the most habitual, trustworthy and useful AI interactions across device ecosystems.