Today, Resemble AI announced its latest breakthrough model, Rapid Voice Clone 2.0, enabling users to generate high-quality voice content with just 20 seconds of audio. This powerful tool allows seamless voice generation, editing, and localization, empowering users to make instant modifications—whether it’s swapping words, fine-tuning tone, or adjusting delivery—without requiring re-recordings.
From crafting dynamic game characters to personalizing narration and translating speech, Rapid Voice Clone 2.0 adapts effortlessly to any scenario. Users can build voice clones with precise control over attributes like gender, emotion, and speed, ensuring more authentic and engaging experiences for audiences.
A Head-to-Head Comparison
To showcase the accuracy of our model, we conducted a direct comparison between Resemble AI and ElevenLabs, using identical 20-second audio clips.
The results were clear: while ElevenLabs produced a reasonable imitation of the original voice in a different language, it failed to retain key prosodic attributes, an issue common across many generative voice AI models.
In contrast, Resemble AI preserved the unique characteristics of the speaker’s voice, ensuring that the clone sounded natural and true to the original accent.
While both models generate similar phrases, Resemble AI captures the speaker’s true accent, making it significantly more natural and lifelike. In a blind survey, 85% of respondents preferred Resemble AI’s voice clones over competitors.
| Accent | Real Speaker (Ground Truth) | Resemble AI (Ours) | ElevenLabs |
|---|---|---|---|
| Australian | |||
| French | |||
| Jamaican |
Accent Integrity in Action
Resemble AI’s latest TTS model leads the market in accent preservation, ensuring that cloned voices retain the unique intonations, rhythm, and speech patterns of the original speaker. Unlike other models that flatten accents or introduce inconsistencies, Resemble AI’s technology maintains regional and cultural authenticity, allowing users to produce highly personalized voice experiences without compromise.
This breakthrough is especially critical for industries requiring authentic voice replication, including entertainment, global media, and customer service. Whether it’s capturing the warmth of a Southern drawl, the cadence of a British accent, or the tonality of an Indian speaker, Resemble AI ensures that voices remain faithful to their original speakers.
From Hollywood to gaming studios, leading brands trust Resemble AI to produce high-quality, diverse voices at scale.
Red Games – Crayola Adventures
Red Games developed Crayola Adventures, an interactive and educational game that transforms traditional storytelling into a choose-your-own-adventure experience. A core element of this project was fostering player creativity and self-expression. By integrating Resemble AI’s dynamic voice cloning technology, the game enables players to personalize their characters and worlds, making each adventure truly their own.
TrueFan – Bollywood Celebrity Voice Cloning
TrueFan, a platform that connects Bollywood celebrities with fans through personalized video messages, leveraged Resemble AI to scale their production from 500 videos per month to 35,000, all while maintaining the authentic voice characteristics of Bollywood stars.
During a Mother’s Day campaign, Resemble AI powered AI-generated celebrity voiceovers, delivering hyper-personalized video messages that deeply resonated with fans. This innovation enabled TrueFan to scale operations while ensuring a highly personalized and emotional connection between celebrities and their audiences.
Resemble AI Joins Google Cloud Marketplace
As of today, Resemble AI is available on Google Cloud Marketplace, making it easier for users to access our Rapid Voice Clone 2.0 solution and explore new possibilities in voice technology. We’re also excited to announce that Resemble AI is now part of the Google Cloud ISV Startup Springboard program.
Bringing Rapid Voice Clone 2.0 to Google Cloud Marketplace will help customers quickly deploy, manage, and grow Rapid Voice Clone 2.0 on Google Cloud’s trusted, global infrastructure,” said Dai Vu, Managing Director, Marketplace & ISV GTM Programs at Google Cloud. “Resemble AI can now securely scale and support customers on their digital transformation journeys.”
With Google Cloud’s infrastructure, Resemble AI is achieving new levels in accuracy and authenticity of synthetic voices. With the integration of Google Cloud’s Gemini models, our AI-powered voices are becoming smarter, offering more natural conversations and better context understanding. This means faster, higher-quality voice production for narration, interactive experiences, and content localization—cutting production time from weeks to just hours while reducing costs.
Ethics and Responsible Use
Looking ahead, we’re committed to refining our technology and expanding its use cases and with our availability on Google Cloud Marketplace we can help more people leverage AI voices. Whether it’s for creating more personalized content, improving the gaming experience, or exploring new industries, Resemble AI remains at the forefront of voice cloning technology.
Resemble AI remains committed to pushing the boundaries of Generative AI while prioritizing AI security. Our dedication to trustworthy AI is evident through advancements like Detect-2B, our state-of-the-art deepfake detection model, and PerTH, our neural watermarking method that embeds imperceptible AI-generated watermarks for enhanced content integrity. As we continue to innovate, we remain steadfast in our mission to deliver authentic, secure, and cutting-edge AI-driven voice technology.
