Post by @thatlev

First we covered GPT-Live-1's launch. Then its tiny win over Grok. Now Gemini has taken the trophy. Artificial Analysis puts Gemini 3.8 Live Extended Thinking at High at 82.6 on its Speech to Speech Index, above GPT-Live-1 with Astra at medium reasoning on 81.5. Gemini also edges ahead on Tau Voice. But Grok still starts speaking sooner in the Big Bench Audio test: 0.70 seconds versus Gemini's 1.35. This is the combined benchmark crown, not a clean sweep. OpenAI barely had time to pose with it #Gemini #GPTLive1 #VoiceAI #ArtificialAnalysis #OpenAI