Qwen Launches Multimodal AI Model Rivaling Google’s Gemini 3.8 at Lower Cost

Qwen’s new multimodal AI model matches Google’s Gemini 3.8 on audio-video benchmarks while providing a lower-cost API alternative, according to The Decoder.

Qwen has introduced its first multimodal AI model, Qwen3.8-Omni-Flash, which demonstrates performance on par with Google's Gemini 3.8 Flash in audio-video benchmark tests. According to The Decoder, Qwen’s new model nearly matches the capabilities of Gemini 3.8 Flash while offering a significantly reduced API cost.

This development marks a notable advancement in AI technology competition, especially in the multimodal domain where audio and video processing are critical. By delivering comparable quality with lower operational expenses, Qwen could influence pricing and adoption trends in AI services.

For Japanese markets, where AI integration in FX, crypto, and equities trading is growing, more affordable and capable multimodal AI models like Qwen3.8-Omni-Flash may accelerate innovation and cost efficiency in data analysis and decision-making tools.

#newsroom

By Mansa Kane

Tokyo Overnight

Japan markets, decoded. Get the brief in your inbox.

No spam. Unsubscribe anytime.