Groundbreaking update enables emotion-aware voice synthesis, multilingual performance, and transformative creative flexibility—all with just a few clicks
SAN FRANCISCO, Dec. 17, 2025 /PRNewswire-PRWeb/ -- Neosapience, creator of Typecast, the platform that empowers expressive, human-quality voice and video creation with AI, today announced the launch of its new voice model, SSFM v3.0, showcasing a major leap forward in voice AI technology designed to serve creators, storytellers, and enterprises. With this release, Typecast fundamentally transforms how synthetic voices express emotion and adapt to context.
The new release introduces automatic emotion recognition, capable of analyzing any script and assigning contextually appropriate emotions to each sentence while enabling cloned voices to perform in styles that go far beyond the original recording. With SSFM v3.0, AI-generated voices are no longer static or flat; they are dynamic and responsive to any story or tone.
The latest model leapfrogs current AI voice technology, which typically requires manual emotion tagging or supports only a limited vocal and emotional range. By contrast, SSFM v3.0's Context-based Automatic Emotion Expression technology can analyze any script—from sales presentations to book readings to breaking news. It automatically generates the appropriate emotional delivery for each sentence without human intervention, making the resulting output indistinguishable from a human voice.
"Our latest Typecast model redefines the boundaries of AI-driven voice," said Taesu Kim, co-founder and CEO of Neosapience. "For the first time, with just a few minutes of human audio we can develop a voice with AI that expands on the original, delivering rich performances well beyond what was recorded. For example, we can create voices that are even more emotive than the speaker is capable of producing on their own. Our vision is to empower creators with new dimensions of expression, opening possibilities that surpass simple replication."
Breakthrough Voice Technology
Unlike traditional voice cloning that can only reproduce recorded tones and delivery styles, SSFM v3.0's Universal Voice Transformation technology can extrapolate from less than five minutes of user-recorded audio samples to create entirely new vocal performances in a wide range of vocal styles and emotions that the original speaker might not be able to achieve naturally on their own. The platform also delivers native-like fluency across six major languages: English, Korean, Japanese, Chinese, Spanish, and Vietnamese, with support for over 30 additional languages. This capability eliminates the accent and tone limitations that have plagued AI voice platforms to this point. By removing this barrier, SSFM v3.0 enables users to instantly scale their content to gain new audiences.
SSFM v3.0 also gives creators unprecedented flexibility through dual emotion control tools through:
- Emotion Design: For the first time, users can create reusable emotions through prompting. Once the prompt is established to define a particular emotion, the user can apply it elsewhere without additional training. This saves time and delivers the right tone over and over again. No one else in the industry has this capability.
- Natural Language Prompting: In addition to Emotion Design, creators can fine-tune intensity with a custom prompt like "make it warmer" or "add more urgency." This makes it simple to adjust delivery for every use.
- Automatic Context Analysis: Typecast's AI also analyzes scripts and assigns appropriate emotions to each sentence based on context, dramatically reducing production time from weeks to minutes.
- Emotion Transfer: Users can also record their own emotional delivery to transfer those feelings to any AI voice.
Together, these tools eliminate the need for labor intensive manual adjustments and allows users to generate polished content with no technical expertise required.
Built for Performance
Powered by a foundation model trained on datasets 10x larger than its predecessor, SSFM v3.0 delivers significantly more realistic and expressive speech. The platform converts natural language emotion descriptions into embedding vectors that guide realistic vocal delivery, representing years of research into the relationship between text, tone, and expression.
"Typecast feels like holding real auditions for the characters in my head," said Gabriel Knight, Film Director & Actor. "I can try different voices, adjust timing, and play with emotions until the story truly comes alive. It's not just a voice tool—it feels like I'm casting, directing, and producing all at once."
The platform supports diverse use cases across industries, including:
- Digital Content Creation: Creators and streaming platforms use Typecast to scale content production without losing quality or emotion.
- Enterprise Integration: Companies embed Typecast's API into their products for conversational AI, virtual assistants, customer training and interactive experiences.
- Brand Communication: Organizations create consistent, professional content for social media and public communications.
- News and Media: Media outlets produce breaking news segments and maintain 24/7 social media presence.
The launch of SSFM v3.0 comes as Neosapience continues strong global growth. The company's average monthly growth rate has exceeded 10% consecutively for the last three years. It now serves more than 2 million users globally.
Typecast SSFM v3.0 is available now. To explore its capabilities, please visit http://typecast.ai.
About Neosapience
Founded in 2017 by former Qualcomm engineers and KAIST graduates, Neosapience develops AI-powered virtual actor technology that makes professional-quality voice and video content creation accessible to creators worldwide. The company's Typecast platform serves over 2 million users globally and has achieved $10 million in annual recurring revenue through its innovative approach to emotional AI voice synthesis. Learn more at www.neosapience.com.
Media Contact
Amber Moore, Moore Communications, 1 5039439381, amber@moorecom2.com
SOURCE Neosapience
Share this article