What Sesame’s New Voice AI App on iOS Is Trying to Be
Sesame’s new voice AI app on iOS is a mobile application that offers four lifelike conversational AI agents designed to act as ongoing, personal voice companions for everyday search, planning, and note‑taking tasks. Now in a free preview across 39 markets, the Sesame app centres on spoken, back‑and‑forth exchanges instead of text prompts, aiming to feel more like talking to a person than using a traditional chatbot or smart speaker. Inside a single session, users can speak to an agent, see live search cards with images, capture notes and summaries, and switch to a text mode when voice is not convenient. The system keeps memory for more personal, tailored exchanges but also offers an incognito mode for conversations that should not be stored. It is both a product test and a broader experiment in whether voice‑first AI interfaces can move from novelty to daily mobile habit.
Four Conversational AI Agents: Maya, Miles, Simone, and Charlie
Instead of one general AI personal assistant, Sesame offers four conversational AI agents—Maya, Miles, Simone, and Charlie—each with distinct voices and personalities. The goal is to give people a choice of conversational style, not only a choice of features. While Sesame does not publish detailed character sheets, the agents are tuned for different tones of dialogue, combining timing, pitch, rhythm, and emotion to feel more natural than a flat text‑to‑speech system. Under the hood, they share the same core voice model and parallel search system, which means any agent can perform live web lookups, summarise results, or keep track of ongoing plans. By tying memory and tone to a specific persona, Sesame is testing whether users will treat a chosen agent as a recurring companion—something they return to repeatedly—rather than as an occasional utility for isolated questions.
Sesame App Features: One Thread for Search, Notes, and Privacy
The Sesame app features are built around an all‑in‑one thread where users can “search, text, and think” without jumping between apps. While you speak, the system runs parallel searches so that live web results and image‑rich search cards can appear before your sentence ends, keeping the conversation fluid. Notes, reminders, and summaries sit inside that same session, turning what starts as a spoken query into a lightweight workflow for planning or research. A discreet text mode lets you continue when speaking aloud is not possible. Memory allows the agents to recall past context for more tailored help, but incognito mode switches to ephemeral conversations that stay off Sesame’s servers and out of long‑term memory. According to TestingCatalog, these agents are optimised “for both speed and thoughtful engagement,” an explicit attempt to balance fast responses with useful depth in everyday use.
From Novelty to Daily Mobile Tool: Can Voice AI Stick?
Sesame’s iOS preview is less about showing off a demo and more about proving that voice AI can handle ordinary phone tasks often done in apps like browsers, notes, or reminders. The company is measuring whether people keep using voice mode for search, planning, and quick task management once the novelty wears off. That means solving several hard problems at once: low latency so first audio comes in under a natural threshold, long‑form context across follow‑up questions, and memory that feels helpful rather than intrusive. If users fall back to tapping in other apps for every follow‑up action, Sesame risks being seen as a polished demo rather than an essential tool. The preview phase, offered free and sometimes behind a waitlist, is effectively a live experiment in turning conversational AI agents into part of the daily iPhone routine.
Where Sesame’s Voice-First Strategy Fits in the AI Assistant Race
Sesame enters a crowded field of conversational AI agents that includes ElevenLabs, OpenAI Realtime, Hume EVI 4, Vapi, and Deepgram, all racing to make speech interfaces feel natural enough for long, frequent use. Sesame’s differentiator is its voice‑first, multi‑agent approach on mobile combined with a longer roadmap: the company is already pointing to intelligent eyewear planned for 2027, suggesting that the iOS voice AI app is a stepping stone toward always‑available assistants. Its earlier 2025 voice demo drew attention for natural timing and turn‑taking, and this broader iPhone rollout moves that research into a daily‑use test at scale. Competition is shifting away from one‑off novelty demos toward systems that can respond quickly, remember context, and support lengthy dialogues. If Sesame can build strong daily habits now, it strengthens the case for voice‑first hardware later.