Qwen3.8-Omni-Flash Challenges Gemini Flash on Multimodal Performance and Price

VERA Brief

AI-generated. Grounded in the article and its cited sources.

Alibaba has released Qwen3.8-Omni-Flash, a new multimodal AI model. This model offers performance comparable to Google's Gemini Flash on audio-video benchmarks, but at a lower API cost.

Key facts

• Alibaba's Qwen3.8-Omni-Flash is a multimodal AI model capable of processing audio and video inputs.
• The model is designed for AI agents and can be used for tasks like vlog editing and clip translation.
• Qwen3.8-Omni-Flash shows performance nearly equivalent to Google's Gemini 3.8 Flash on audio-video benchmarks.
• The API cost for Qwen3.8-Omni-Flash is substantially lower than that of Gemini Flash.
• This release allows for verification of multimodal AI system performance and economic viability.