Gemini 2.5 Pro Gets Even Better, ElevenLabs V3 Leaves Us Speechless, Pepsi Unveils The FizzBot!

AI News - June 05

Gemini 2.5 gets another performance boost ElevenLabs takes text-to-speech to a whole new level the new Higgsfield update speaks for itself and Pepsi unveils its FizzBot Here's today's AI news. Google are rolling out another update to Gemini 2.5 Pro, which boosts its performance across coding, math, science and reasoning.

[00:20.7]
It now scores almost 22% in humanity's last exam. It seems like Google are just going to keep raising the bar with 2.5 Pro. ElevenLabs have introduced their latest model Eleven V3 and they claim it's the most expressive text to speech model ever.

[00:36.4]
From the demonstrations we've seen so far, it would be very hard to argue with that. The range of expression and the control seem second to none. This really does take text to speech to a whole new level. But elevenlabs aren't the only ones who are pushing the boundaries with AI speech.

[00:51.4]
Today, Higgs Field released Higgsfield Speak that allows you to add a script to a video and have the words spoken via an avatar, with Veo 3 adding sound and dialogue to their output. It seems that simply creating visuals from a video model is no longer enough. And finally, Pepsi has partnered with AGIBOT to create a commercial partnership that literally no one was asking for.

[01:12.7]
They unveiled a Pepsi branded humanoid robot called Fizzbot and it shared the stage with David Beckham at the launch event. They say it's for promotional purposes, but we're not sure this is going to convince anyone to drink more Pepsi.