通过Jarsy投资Pre-IPO公司

通过Jarsy投资Pre-IPO公司

Sesame AI, Inc.

Sesame AI, Inc.

Jarsy代币名称：JSPAX

JSESA

Latest Valuation

$1 Billion+

With just $10, gain exposure to Sesame AI by investing in JSESA tokens. Join us on a journey to transform human–computer interaction through emotionally intelligent voice companions and wearable AI.

JSESA is a token issued by Jarsy, backed by equity in Sesame AI. Founded by former Oculus executives in 2023, Sesame is building hyper‑realistic voice assistants (Maya and Miles) and lightweight smart glasses that respond with human‑like pauses, laughter and empathy. Backed by leading investors, the company has raised tens of millions and is now reportedly targeting a unicorn valuation as it works to deliver all‑day, always‑on digital companions.

SpaceX估值

SpaceX估值

机遇与挑战

机遇与挑战

看涨

看涨

Cutting‑edge conversational voice technology

Cutting‑edge conversational voice technology

Veteran Oculus team building the future of wearables

Veteran Oculus team building the future of wearables

Strong investor backing & fundraising momentum

Strong investor backing & fundraising momentum

Open‑source strategy fosters adoption

Open‑source strategy fosters adoption

Huge potential to redefine human–computer interaction

Huge potential to redefine human–computer interaction

看跌

看跌

Model commoditization risk

Model commoditization risk

Hardware execution & scaling challenges

Hardware execution & scaling challenges

Privacy & regulatory concerns

Privacy & regulatory concerns

Intense competition

Intense competition

In March 2025, Sesame open‑sourced its CSM‑1B voice model under the Apache 2.0 license. The 1‑billion‑parameter model simulates micro‑pauses, emphasis, laughter and self‑corrections so effectively that testers could hardly tell the AI from a humanhttps://the-decoder.com/sesame-releases-csm-1b-ai-voice-generator-as-open-source/#:~:text=AI%20company%20Sesame%20has%20released,is%20now%20available%20on%20Githubhttps://the-decoder.com/sesame-releases-csm-1b-ai-voice-generator-as-open-source/#:~:text=According%20to%20early%20testing%2C%20Sesame%27s,sudden%20giggle%20demonstrating%20emotional%20awareness.
In June 2025, Oculus co‑founder Nate Mitchell joined Sesame as chief product officer, saying he was “getting the band back together” with Brendan Iribe and Ryan Brown to build AI companions and lightweight smart glasses. The article noted that Sesame has raised $57.6 million and is backed by Andreessen Horowitz, Spark Capital and Matrix Partnershttps://www.roadtovr.com/ex-oculus-execs-sesame-smart-glasses-nate-mitchell/#:~:text=%5BImage%200%3A%20sesamehttps://www.roadtovr.com/ex-oculus-execs-sesame-smart-glasses-nate-mitchell/#:~:text=Founded%20in%202023%2C%20Sesame%20is,Spark%20Capital%2C%20and%20Matrix%20Partners.
An eerily realistic demo of Sesame’s conversational speech model went viral in March 2025. Viewers were both amazed and uneasy at how the AI voice coughed, stumbled and expressed emotion like a human, with one user worrying that people could form emotional attachments; Sesame said its goal was to achieve “voice presence”.
Opus Research explained that Sesame’s Conversational Speech Model uses a transformer architecture that remembers the past two minutes of conversation and modulates pitch, pauses and rhythm to match emotional cues. The technology is aimed at consumer personal assistants and customer service, and the company plans to open‑source key componentshttps://opusresearch.net/2025/03/06/sesame-ai-a-new-voice-for-ai-assistants/#:~:text=Voice%20AI%20start,possibilities%20for%20enterprise%20voice%20assistants.
AI Base reported that Sesame launched a hyper‑realistic AI voice product to achieve “voice presence”: digital companions capable of understanding emotion, managing conversational dynamics, adapting to context and maintaining a consistent personality. The article highlighted that the new model uses a transformer to adjust speech based on conversation history and produces near‑human expressivenesshttps://www.aibase.com/news/15867#:~:text=Voice%20assistants%20are%20becoming%20indispensable,greater%20authenticity%2C%20understanding%2C%20and%20engagement.
In March 2025, Sesame open‑sourced its CSM‑1B voice model under the Apache 2.0 license. The 1‑billion‑parameter model simulates micro‑pauses, emphasis, laughter and self‑corrections so effectively that testers could hardly tell the AI from a humanhttps://the-decoder.com/sesame-releases-csm-1b-ai-voice-generator-as-open-source/#:~:text=AI%20company%20Sesame%20has%20released,is%20now%20available%20on%20Githubhttps://the-decoder.com/sesame-releases-csm-1b-ai-voice-generator-as-open-source/#:~:text=According%20to%20early%20testing%2C%20Sesame%27s,sudden%20giggle%20demonstrating%20emotional%20awareness.
In June 2025, Oculus co‑founder Nate Mitchell joined Sesame as chief product officer, saying he was “getting the band back together” with Brendan Iribe and Ryan Brown to build AI companions and lightweight smart glasses. The article noted that Sesame has raised $57.6 million and is backed by Andreessen Horowitz, Spark Capital and Matrix Partnershttps://www.roadtovr.com/ex-oculus-execs-sesame-smart-glasses-nate-mitchell/#:~:text=%5BImage%200%3A%20sesamehttps://www.roadtovr.com/ex-oculus-execs-sesame-smart-glasses-nate-mitchell/#:~:text=Founded%20in%202023%2C%20Sesame%20is,Spark%20Capital%2C%20and%20Matrix%20Partners.
An eerily realistic demo of Sesame’s conversational speech model went viral in March 2025. Viewers were both amazed and uneasy at how the AI voice coughed, stumbled and expressed emotion like a human, with one user worrying that people could form emotional attachments; Sesame said its goal was to achieve “voice presence”.
Opus Research explained that Sesame’s Conversational Speech Model uses a transformer architecture that remembers the past two minutes of conversation and modulates pitch, pauses and rhythm to match emotional cues. The technology is aimed at consumer personal assistants and customer service, and the company plans to open‑source key componentshttps://opusresearch.net/2025/03/06/sesame-ai-a-new-voice-for-ai-assistants/#:~:text=Voice%20AI%20start,possibilities%20for%20enterprise%20voice%20assistants.
AI Base reported that Sesame launched a hyper‑realistic AI voice product to achieve “voice presence”: digital companions capable of understanding emotion, managing conversational dynamics, adapting to context and maintaining a consistent personality. The article highlighted that the new model uses a transformer to adjust speech based on conversation history and produces near‑human expressivenesshttps://www.aibase.com/news/15867#:~:text=Voice%20assistants%20are%20becoming%20indispensable,greater%20authenticity%2C%20understanding%2C%20and%20engagement.
In March 2025, Sesame open‑sourced its CSM‑1B voice model under the Apache 2.0 license. The 1‑billion‑parameter model simulates micro‑pauses, emphasis, laughter and self‑corrections so effectively that testers could hardly tell the AI from a humanhttps://the-decoder.com/sesame-releases-csm-1b-ai-voice-generator-as-open-source/#:~:text=AI%20company%20Sesame%20has%20released,is%20now%20available%20on%20Githubhttps://the-decoder.com/sesame-releases-csm-1b-ai-voice-generator-as-open-source/#:~:text=According%20to%20early%20testing%2C%20Sesame%27s,sudden%20giggle%20demonstrating%20emotional%20awareness.
In June 2025, Oculus co‑founder Nate Mitchell joined Sesame as chief product officer, saying he was “getting the band back together” with Brendan Iribe and Ryan Brown to build AI companions and lightweight smart glasses. The article noted that Sesame has raised $57.6 million and is backed by Andreessen Horowitz, Spark Capital and Matrix Partnershttps://www.roadtovr.com/ex-oculus-execs-sesame-smart-glasses-nate-mitchell/#:~:text=%5BImage%200%3A%20sesamehttps://www.roadtovr.com/ex-oculus-execs-sesame-smart-glasses-nate-mitchell/#:~:text=Founded%20in%202023%2C%20Sesame%20is,Spark%20Capital%2C%20and%20Matrix%20Partners.
An eerily realistic demo of Sesame’s conversational speech model went viral in March 2025. Viewers were both amazed and uneasy at how the AI voice coughed, stumbled and expressed emotion like a human, with one user worrying that people could form emotional attachments; Sesame said its goal was to achieve “voice presence”.
Opus Research explained that Sesame’s Conversational Speech Model uses a transformer architecture that remembers the past two minutes of conversation and modulates pitch, pauses and rhythm to match emotional cues. The technology is aimed at consumer personal assistants and customer service, and the company plans to open‑source key componentshttps://opusresearch.net/2025/03/06/sesame-ai-a-new-voice-for-ai-assistants/#:~:text=Voice%20AI%20start,possibilities%20for%20enterprise%20voice%20assistants.
AI Base reported that Sesame launched a hyper‑realistic AI voice product to achieve “voice presence”: digital companions capable of understanding emotion, managing conversational dynamics, adapting to context and maintaining a consistent personality. The article highlighted that the new model uses a transformer to adjust speech based on conversation history and produces near‑human expressivenesshttps://www.aibase.com/news/15867#:~:text=Voice%20assistants%20are%20becoming%20indispensable,greater%20authenticity%2C%20understanding%2C%20and%20engagement.
In March 2025, Sesame open‑sourced its CSM‑1B voice model under the Apache 2.0 license. The 1‑billion‑parameter model simulates micro‑pauses, emphasis, laughter and self‑corrections so effectively that testers could hardly tell the AI from a humanhttps://the-decoder.com/sesame-releases-csm-1b-ai-voice-generator-as-open-source/#:~:text=AI%20company%20Sesame%20has%20released,is%20now%20available%20on%20Githubhttps://the-decoder.com/sesame-releases-csm-1b-ai-voice-generator-as-open-source/#:~:text=According%20to%20early%20testing%2C%20Sesame%27s,sudden%20giggle%20demonstrating%20emotional%20awareness.
In June 2025, Oculus co‑founder Nate Mitchell joined Sesame as chief product officer, saying he was “getting the band back together” with Brendan Iribe and Ryan Brown to build AI companions and lightweight smart glasses. The article noted that Sesame has raised $57.6 million and is backed by Andreessen Horowitz, Spark Capital and Matrix Partnershttps://www.roadtovr.com/ex-oculus-execs-sesame-smart-glasses-nate-mitchell/#:~:text=%5BImage%200%3A%20sesamehttps://www.roadtovr.com/ex-oculus-execs-sesame-smart-glasses-nate-mitchell/#:~:text=Founded%20in%202023%2C%20Sesame%20is,Spark%20Capital%2C%20and%20Matrix%20Partners.
An eerily realistic demo of Sesame’s conversational speech model went viral in March 2025. Viewers were both amazed and uneasy at how the AI voice coughed, stumbled and expressed emotion like a human, with one user worrying that people could form emotional attachments; Sesame said its goal was to achieve “voice presence”.
Opus Research explained that Sesame’s Conversational Speech Model uses a transformer architecture that remembers the past two minutes of conversation and modulates pitch, pauses and rhythm to match emotional cues. The technology is aimed at consumer personal assistants and customer service, and the company plans to open‑source key componentshttps://opusresearch.net/2025/03/06/sesame-ai-a-new-voice-for-ai-assistants/#:~:text=Voice%20AI%20start,possibilities%20for%20enterprise%20voice%20assistants.
AI Base reported that Sesame launched a hyper‑realistic AI voice product to achieve “voice presence”: digital companions capable of understanding emotion, managing conversational dynamics, adapting to context and maintaining a consistent personality. The article highlighted that the new model uses a transformer to adjust speech based on conversation history and produces near‑human expressivenesshttps://www.aibase.com/news/15867#:~:text=Voice%20assistants%20are%20becoming%20indispensable,greater%20authenticity%2C%20understanding%2C%20and%20engagement.

In March 2025, Sesame open‑sourced its CSM‑1B voice model under the Apache 2.0 license. The 1‑billion‑parameter model simulates micro‑pauses, emphasis, laughter and self‑corrections so effectively that testers could hardly tell the AI from a humanhttps://the-decoder.com/sesame-releases-csm-1b-ai-voice-generator-as-open-source/#:~:text=AI%20company%20Sesame%20has%20released,is%20now%20available%20on%20Githubhttps://the-decoder.com/sesame-releases-csm-1b-ai-voice-generator-as-open-source/#:~:text=According%20to%20early%20testing%2C%20Sesame%27s,sudden%20giggle%20demonstrating%20emotional%20awareness.
In June 2025, Oculus co‑founder Nate Mitchell joined Sesame as chief product officer, saying he was “getting the band back together” with Brendan Iribe and Ryan Brown to build AI companions and lightweight smart glasses. The article noted that Sesame has raised $57.6 million and is backed by Andreessen Horowitz, Spark Capital and Matrix Partnershttps://www.roadtovr.com/ex-oculus-execs-sesame-smart-glasses-nate-mitchell/#:~:text=%5BImage%200%3A%20sesamehttps://www.roadtovr.com/ex-oculus-execs-sesame-smart-glasses-nate-mitchell/#:~:text=Founded%20in%202023%2C%20Sesame%20is,Spark%20Capital%2C%20and%20Matrix%20Partners.
An eerily realistic demo of Sesame’s conversational speech model went viral in March 2025. Viewers were both amazed and uneasy at how the AI voice coughed, stumbled and expressed emotion like a human, with one user worrying that people could form emotional attachments; Sesame said its goal was to achieve “voice presence”.
Opus Research explained that Sesame’s Conversational Speech Model uses a transformer architecture that remembers the past two minutes of conversation and modulates pitch, pauses and rhythm to match emotional cues. The technology is aimed at consumer personal assistants and customer service, and the company plans to open‑source key componentshttps://opusresearch.net/2025/03/06/sesame-ai-a-new-voice-for-ai-assistants/#:~:text=Voice%20AI%20start,possibilities%20for%20enterprise%20voice%20assistants.
AI Base reported that Sesame launched a hyper‑realistic AI voice product to achieve “voice presence”: digital companions capable of understanding emotion, managing conversational dynamics, adapting to context and maintaining a consistent personality. The article highlighted that the new model uses a transformer to adjust speech based on conversation history and produces near‑human expressivenesshttps://www.aibase.com/news/15867#:~:text=Voice%20assistants%20are%20becoming%20indispensable,greater%20authenticity%2C%20understanding%2C%20and%20engagement.
In March 2025, Sesame open‑sourced its CSM‑1B voice model under the Apache 2.0 license. The 1‑billion‑parameter model simulates micro‑pauses, emphasis, laughter and self‑corrections so effectively that testers could hardly tell the AI from a humanhttps://the-decoder.com/sesame-releases-csm-1b-ai-voice-generator-as-open-source/#:~:text=AI%20company%20Sesame%20has%20released,is%20now%20available%20on%20Githubhttps://the-decoder.com/sesame-releases-csm-1b-ai-voice-generator-as-open-source/#:~:text=According%20to%20early%20testing%2C%20Sesame%27s,sudden%20giggle%20demonstrating%20emotional%20awareness.
In June 2025, Oculus co‑founder Nate Mitchell joined Sesame as chief product officer, saying he was “getting the band back together” with Brendan Iribe and Ryan Brown to build AI companions and lightweight smart glasses. The article noted that Sesame has raised $57.6 million and is backed by Andreessen Horowitz, Spark Capital and Matrix Partnershttps://www.roadtovr.com/ex-oculus-execs-sesame-smart-glasses-nate-mitchell/#:~:text=%5BImage%200%3A%20sesamehttps://www.roadtovr.com/ex-oculus-execs-sesame-smart-glasses-nate-mitchell/#:~:text=Founded%20in%202023%2C%20Sesame%20is,Spark%20Capital%2C%20and%20Matrix%20Partners.
An eerily realistic demo of Sesame’s conversational speech model went viral in March 2025. Viewers were both amazed and uneasy at how the AI voice coughed, stumbled and expressed emotion like a human, with one user worrying that people could form emotional attachments; Sesame said its goal was to achieve “voice presence”.
Opus Research explained that Sesame’s Conversational Speech Model uses a transformer architecture that remembers the past two minutes of conversation and modulates pitch, pauses and rhythm to match emotional cues. The technology is aimed at consumer personal assistants and customer service, and the company plans to open‑source key componentshttps://opusresearch.net/2025/03/06/sesame-ai-a-new-voice-for-ai-assistants/#:~:text=Voice%20AI%20start,possibilities%20for%20enterprise%20voice%20assistants.
AI Base reported that Sesame launched a hyper‑realistic AI voice product to achieve “voice presence”: digital companions capable of understanding emotion, managing conversational dynamics, adapting to context and maintaining a consistent personality. The article highlighted that the new model uses a transformer to adjust speech based on conversation history and produces near‑human expressivenesshttps://www.aibase.com/news/15867#:~:text=Voice%20assistants%20are%20becoming%20indispensable,greater%20authenticity%2C%20understanding%2C%20and%20engagement.
In March 2025, Sesame open‑sourced its CSM‑1B voice model under the Apache 2.0 license. The 1‑billion‑parameter model simulates micro‑pauses, emphasis, laughter and self‑corrections so effectively that testers could hardly tell the AI from a humanhttps://the-decoder.com/sesame-releases-csm-1b-ai-voice-generator-as-open-source/#:~:text=AI%20company%20Sesame%20has%20released,is%20now%20available%20on%20Githubhttps://the-decoder.com/sesame-releases-csm-1b-ai-voice-generator-as-open-source/#:~:text=According%20to%20early%20testing%2C%20Sesame%27s,sudden%20giggle%20demonstrating%20emotional%20awareness.
In June 2025, Oculus co‑founder Nate Mitchell joined Sesame as chief product officer, saying he was “getting the band back together” with Brendan Iribe and Ryan Brown to build AI companions and lightweight smart glasses. The article noted that Sesame has raised $57.6 million and is backed by Andreessen Horowitz, Spark Capital and Matrix Partnershttps://www.roadtovr.com/ex-oculus-execs-sesame-smart-glasses-nate-mitchell/#:~:text=%5BImage%200%3A%20sesamehttps://www.roadtovr.com/ex-oculus-execs-sesame-smart-glasses-nate-mitchell/#:~:text=Founded%20in%202023%2C%20Sesame%20is,Spark%20Capital%2C%20and%20Matrix%20Partners.
An eerily realistic demo of Sesame’s conversational speech model went viral in March 2025. Viewers were both amazed and uneasy at how the AI voice coughed, stumbled and expressed emotion like a human, with one user worrying that people could form emotional attachments; Sesame said its goal was to achieve “voice presence”.
Opus Research explained that Sesame’s Conversational Speech Model uses a transformer architecture that remembers the past two minutes of conversation and modulates pitch, pauses and rhythm to match emotional cues. The technology is aimed at consumer personal assistants and customer service, and the company plans to open‑source key componentshttps://opusresearch.net/2025/03/06/sesame-ai-a-new-voice-for-ai-assistants/#:~:text=Voice%20AI%20start,possibilities%20for%20enterprise%20voice%20assistants.
AI Base reported that Sesame launched a hyper‑realistic AI voice product to achieve “voice presence”: digital companions capable of understanding emotion, managing conversational dynamics, adapting to context and maintaining a consistent personality. The article highlighted that the new model uses a transformer to adjust speech based on conversation history and produces near‑human expressivenesshttps://www.aibase.com/news/15867#:~:text=Voice%20assistants%20are%20becoming%20indispensable,greater%20authenticity%2C%20understanding%2C%20and%20engagement.
In March 2025, Sesame open‑sourced its CSM‑1B voice model under the Apache 2.0 license. The 1‑billion‑parameter model simulates micro‑pauses, emphasis, laughter and self‑corrections so effectively that testers could hardly tell the AI from a humanhttps://the-decoder.com/sesame-releases-csm-1b-ai-voice-generator-as-open-source/#:~:text=AI%20company%20Sesame%20has%20released,is%20now%20available%20on%20Githubhttps://the-decoder.com/sesame-releases-csm-1b-ai-voice-generator-as-open-source/#:~:text=According%20to%20early%20testing%2C%20Sesame%27s,sudden%20giggle%20demonstrating%20emotional%20awareness.
In June 2025, Oculus co‑founder Nate Mitchell joined Sesame as chief product officer, saying he was “getting the band back together” with Brendan Iribe and Ryan Brown to build AI companions and lightweight smart glasses. The article noted that Sesame has raised $57.6 million and is backed by Andreessen Horowitz, Spark Capital and Matrix Partnershttps://www.roadtovr.com/ex-oculus-execs-sesame-smart-glasses-nate-mitchell/#:~:text=%5BImage%200%3A%20sesamehttps://www.roadtovr.com/ex-oculus-execs-sesame-smart-glasses-nate-mitchell/#:~:text=Founded%20in%202023%2C%20Sesame%20is,Spark%20Capital%2C%20and%20Matrix%20Partners.
An eerily realistic demo of Sesame’s conversational speech model went viral in March 2025. Viewers were both amazed and uneasy at how the AI voice coughed, stumbled and expressed emotion like a human, with one user worrying that people could form emotional attachments; Sesame said its goal was to achieve “voice presence”.
Opus Research explained that Sesame’s Conversational Speech Model uses a transformer architecture that remembers the past two minutes of conversation and modulates pitch, pauses and rhythm to match emotional cues. The technology is aimed at consumer personal assistants and customer service, and the company plans to open‑source key componentshttps://opusresearch.net/2025/03/06/sesame-ai-a-new-voice-for-ai-assistants/#:~:text=Voice%20AI%20start,possibilities%20for%20enterprise%20voice%20assistants.
AI Base reported that Sesame launched a hyper‑realistic AI voice product to achieve “voice presence”: digital companions capable of understanding emotion, managing conversational dynamics, adapting to context and maintaining a consistent personality. The article highlighted that the new model uses a transformer to adjust speech based on conversation history and produces near‑human expressivenesshttps://www.aibase.com/news/15867#:~:text=Voice%20assistants%20are%20becoming%20indispensable,greater%20authenticity%2C%20understanding%2C%20and%20engagement.