# Sesame

> Voice AI company whose conversational agents run in iOS and Android apps, built on its Conversational Speech Model.

Sesame is a voice AI company that makes conversational agents for iOS and Android. Sesame agents hold spoken conversations to help users get things done, talk ideas through and explore questions. The agents are built on the Sesame Conversational Speech Model, which uses dialogue history to shape tone and timing. Users download the app to begin.

**Best for:** People who want to interact with an AI agent mainly by natural speech on a phone, and researchers following conversational speech generation.

- Official site: https://www.sesame.com
- Listed in: [Voice Assistants & Speech Recognition](https://trendhub.net/ai-tools/category/voice-assistants), [Text to Speech](https://trendhub.net/ai-tools/category/text-to-speech)
- Last checked: 2026-10-09 (facts taken from sesame.com)
- Source page: https://trendhub.net/ai-tools/sesame-conversational-voice-ai-agents

## Key facts

- Platforms: iOS, Android
- Made by: Sesame

## What is Sesame?

Sesame is a company developing voice agents intended for natural, spoken conversation, together with devices to carry them. Its agents are offered in apps for iOS and Android, and the site says eyewear is planned. The stated goal is what Sesame calls voice presence, which it breaks into four parts: recognising emotional context, handling the timing of real dialogue including pauses and interruptions, adapting tone to the situation and keeping a stable personality. Underneath is the Conversational Speech Model, a transformer-based system that uses the history of a conversation, both text and audio, when generating speech, and works in a single stage. Sesame has published research on the model, offers a research preview and has said its models would be released under an Apache 2.0 license. The company lists offices in New York, San Francisco and Bellevue.

## Key features

- **Voice Agents**: Hold spoken conversations with users to help with tasks, talk through ideas and explore topics.
- **Mobile Apps**: Delivers the agents through downloadable apps for iOS and Android.
- **Conversational Speech Model**: Generates speech with a single-stage transformer model that conditions on prior turns of the dialogue.
- **Expressive Delivery**: Aims for speech with natural pacing, emphasis and tone suited to the emotional context of the exchange.
- **Open Research**: Publishes technical write-ups and evaluation results, with models announced under an Apache 2.0 license.
- **Planned Eyewear**: Describes eyewear as a forthcoming way to use the agents, alongside the company's wider hardware work.

## Use cases

- **Thinking Out Loud**: Talk a problem through with an agent instead of typing it.
- **Following Curiosity**: Ask follow-up questions on a subject in a spoken back-and-forth.
- **Everyday Help by Voice**: Use a phone-based agent for day-to-day tasks without a keyboard.
- **Studying Speech Generation**: Read the published research on contextual speech synthesis and its evaluation.
- **Trying the Research Preview**: Test the conversational voice demo that Sesame makes available.

## Things to know

- **English-centred speech model**: The research post says CSM was trained primarily on English and performs poorly in other languages for now.
- **Acknowledged gap to human speech**: Sesame's own evaluation found listeners still preferred human recordings when conversational context was provided.
- **Eyewear is not yet available**: The home page lists eyewear as coming soon; only the mobile apps are offered today.
- **Limited public detail**: The site does not state prices or plan terms for the apps.

## FAQs

### What is Sesame?

Sesame is a voice AI company that builds conversational agents and the hardware to use them; its agents are available in iOS and Android apps.

### What is Sesame used for?

Sesame agents are used for spoken conversation: getting help with tasks, talking things through and exploring topics by voice.

### Does Sesame work on mobile?

Yes. Sesame is available on iOS and Android, and the company says eyewear is coming.

### What is the Conversational Speech Model?

The Conversational Speech Model, or CSM, is Sesame's transformer-based speech generation model, which uses conversation history to produce more natural and coherent speech.

### Is Sesame's model open source?

Sesame says it plans to open-source key parts of its research and to release its models with an Apache 2.0 license.

### Which languages does Sesame's speech model support?

Sesame's research post says CSM was trained mainly on English data and does not yet perform well in other languages, with wider language support planned.

### Who leads Sesame?

Sesame names five leaders on its team page: Brendan Iribe, Ryan Brown, Ankit Kumar, Nate Mitchell and Angela Gayles.

---

Source: https://trendhub.net/ai-tools/sesame-conversational-voice-ai-agents (trendhub)