Skip to content
← Tools / Skills
ToolSesame/10 trustVerified
Sesame logo

Sesame

Eerily lifelike conversational voice agents, now generally available on mobile.

Visit official site →

About

Sesame is a San Francisco startup co-founded by former Oculus CEO Brendan Iribe that builds conversational voice AI. Its research preview in early 2025 introduced the voices Maya and Miles, which drew more than a million users in the first weeks and logged over five million minutes of conversation. The company raised a $250 million Series B led by Sequoia Capital and Spark Capital in October 2025.

In October 2026 the product reached general availability on iOS and Android with four agents: Maya, Miles, Simone and Charlie. Each has its own voice, personality, memory and permission scope, plus a separate computer environment for web research, tool use and long-running tasks. The app supports natural interruption, persistent memory, scheduled workflows, connected apps, a quiet texting mode and an incognito mode. Sesame is also building AI smart glasses targeted for 2027.

The company open-sourced CSM-1B, the 1-billion-parameter base model behind Maya, under Apache 2.0, while the production stack remains proprietary and details on pricing and benchmarks are still thin.

What it can do

  • Four voice agents with distinct personalities and memories
  • Natural interruption and low-latency conversation
  • Web research woven into live speech
  • Separate computer environment per agent for tasks
  • Scheduled workflows and connected apps
  • Texting mode and incognito mode

Setup

  1. Download the Sesame app on iOS or Android.
  2. Create an account and pick an agent (Maya, Miles, Simone or Charlie).
  3. Start talking; connect apps and enable memory for ongoing tasks.

Prerequisites

A smartphone; a Sesame account.

Use cases

voice-assistantconversational-aihands-free-productivityai-companions

Works with

Trust score — 8/10

/10 trustVerified

Scored 8/10 (verified tier, metrics-v1). Evidence: $250M Series B (Sequoia, Spark) Oct 2025 reported by multiple outlets; 1M+ preview users and 5M+ conversation minutes per company/Sequoia via AlphaSignal; GA on iOS/Android Oct 2026 with four agents (AlphaSignal). Strengths: exceptional voice realism, strong funding, shipping hardware-adjacent product. Docked for: (1) production model pricing and benchmarks undisclosed; (2) long-term viability tied to unshipped glasses. License: proprietary (CSM-1B base model Apache-2.0).

How the trust score is built

Full methodology →
Adoption · stars/downloads, log-scaled vs peers
9.0/10
Maintenance · recency, releases, issue load
9.0/10
Openness · license + docs
6.0/10
Manual review · sanity read + red-flag check
8.0/10
  • · $250M Series B (Sequoia, Spark) Oct 2025 reported by multiple outlets (dataconomy, startupstorymedia)
  • · 1M+ preview users, 5M+ conversation minutes per company/Sequoia via AlphaSignal
  • · GA Oct 2026 on iOS/Android with Maya, Miles, Simone, Charlie; separate compute per agent (AlphaSignal)
  • · Production pricing/benchmarks undisclosed; viability tied to 2027 smart glasses

Sources

Every fact above was checked against these sources. Where a claim comes from the provider and couldn't be independently verified, the text says so.