A new form of eyewear

At Sesame, we’re making all-day glasses that sit alongside the best eyewear in the world, with an assistant you enjoy talking to.

This morning, we’re sharing a fresh preview of our assistant, powered by a new voice model, available in the Sesame app for iOS and Android.

We’re designing the glasses and assistant as one product. Today, intelligence still lives behind a screen. To become present, it needs new hardware built around it.

Eyewear is the right form factor. We wear glasses all day, and take pride in our favorite frames. They’re also perfect for voice.

We think of our roadmap as three parts:

• Frames you love to wear
• Intelligence & assistance you want to use
• Characters you like to talk to

Frames you love to wear

A great pair of glasses is the sum of countless details.

The weight. The feel of the hinges. The fit. The materials. The polish. Each may be nearly invisible on its own, but together they create a pair of glasses that simply feels right – the pair you reach for every day. The pair you love to wear.

Yesterday, we announced that our collection will be Made in Japan by makers who’ve spent decades perfecting their craft.

Great frames are shaped, polished, and fitted by hand over days. There are no shortcuts.

We’ll share more on the collection in the coming months.

Intelligence & assistance you want to use

There’s no shortage of agents you can send a task to. We’re taking a different approach: designing our assistants to be the operating system for the Sesame glasses. One that’s voice first, and as natural to use as talking to a friend.

Our assistants can use tools and help manage other agents on your behalf. You can instantly talk to them anytime, hands-free. And with the October release, each agent is now smarter, with access to its own computer.

You can ask them to plan your next adventure, riff with you on a personal project, or rabbit hole on your favorite hobby. They’ll go off and come back with an update when it’s ready.

1/3

We’ve been using Maya, Miles, Charlie, and Simone to develop this release, helping orchestrate other agents like Devin, Claude Code, and Codex. We do this over voice on walks, commuting, and often at our desks. It’s a new superpower. Now, you can try it yourself.

Our assistants can also connect to other services you use every day. We’re starting with a few Google apps, with more coming soon. You can connect your own MCPs as well.

We’re adding developer tooling so you can make your own Sesame apps later this year. Let us know what you’d like to see.

Characters you like to talk to

When these characters are with you all day, the real test is whether you want to talk to them. We bet early on personality with Maya and Miles.

We develop our characters the way a writers’ room develops characters for film or TV. Writers, producers, and actors work side by side with our research team on every detail: who each character is, what they care about, and how they see the world. We craft voice and personality as one.

Behind each character sits a body of creative work – scripts, character bibles, and scenes – that define them and tell their story. We write, test, and rewrite until each one feels like someone you actually know.

We want there to be a character for everyone. Now their distinct personalities come through more clearly.

We’ve reduced sycophancy and cut down on repetitive phrases. When you’re thinking out loud, building something, or chasing an idea, you want someone who will work through it with you and tell you when you’re wrong.

We’ve also updated our Conversational Speech Model to improve expressivity and pronunciation. This makes a huge difference, especially over longer conversations.

What’s next

We have more updates coming later this year, including better memory, new characters, and new developer tools.

Thank you for all the feedback. More to come.