- Google has announced “Gemini 3.1 Flash Live,” a next-generation voice-first AI model evolved with top-tier real-time conversational capabilities.
- An AI model specialized for dialogue, featuring improved audio understanding compared to its predecessor, the “Gemini 2.5 Flash Native Audio.”
- It has begun rollout as the base AI model for “Gemini Live” and “Search Live.”
On Thursday, March 26, 2026, Google announced “Gemini 3.1 Flash Live,” its next-generation voice-first AI model that has evolved to feature top-tier real-time conversational capabilities.
“Gemini 3.1 Flash Live” is a conversation-specialized AI model with enhanced audio comprehension capabilities compared to the previous “Gemini 2.5 Flash Native Audio” model. It has evolved to more effectively recognize acoustic nuances such as pitch and pace, allowing it to handle complex tasks in noisy environments and respond accurately to emotional expressions, such as user frustration or confusion.
“Gemini 3.1 Flash Live” is being rolled out to developers and has also begun implementation as the base AI model for “Gemini Live” and “Search Live,” meaning general users can now experience this next-generation voice-first AI model through these services.
Source: Google
コメントを残す