Alibaba Cloud unveiled Qwen3.8-Omni-Flash, a 1-million-token omni-modal model optimized for real-time audio-video perception, low-latency voice, and agentic tool use with a 60% inference cost reduction. Concurrently, Alibaba announced an in-cabin AI partnership with BYD.
Key Takeaways
- ✓Supports 1M token context for low-latency full-duplex omni-modal conversational streaming
- ✓Slashes multimodal and audio inference expenses by more than 60%
- ✓Powers BYD next-generation smart cockpit with autonomous in-cabin task orchestration
Discussion & Comments
0Sign in to join the discussion
Connect with AI developers to exchange benchmark insights.