May 2025 Summaries

Anthropic's new Claude Sonnet 4 model, now available on the ElevenLabs platform, offers significant advancements in conversational AI, enabling more sophisticated, intuitive, and responsive voice experiences. Designed to handle complex, multi-turn interactions, Sonnet 4 excels in instruction following, maintaining context over longer dialogues, and reliably using tools to interact with external systems, making it ideal for developing dynamic and efficient voice agents. It addresses key challenges in voice AI by combining speed and intelligence, allowing for real-time conversations without compromising on performance. This facilitates the creation of truly conversational applications capable of managing real-world tasks smoothly, enhancing user engagement, and reducing the need for elaborate workarounds, ultimately leading to more natural, personalized user experiences. The model is easily accessible through the ElevenLabs platform, enabling developers to deploy it with minimal effort and start building advanced voice agents immediately.

ElevenLabs has launched Conversational AI 2.0, an advanced platform aimed at creating sophisticated voice agents that offer seamless, natural interactions. This version introduces significant improvements such as state-of-the-art turn-taking models for smoother dialogue, integrated automatic language detection for multilingual communication, and Retrieval-Augmented Generation (RAG) for accessing external knowledge sources with minimal latency and maximum privacy. The platform also supports multimodality, enabling agents to operate across voice and text simultaneously, and includes batch call capabilities for efficient outbound voice communications. Built with enterprise readiness in mind, it ensures HIPAA compliance, enhanced security, and optional EU data residency, making it suitable for critical business functions and a wide range of applications. Conversational AI 2.0 represents a substantial advancement over its predecessor, underscoring ElevenLabs' commitment to rapid innovation and enterprise-grade reliability.

Ryan Morrison describes the development of an AI-powered tool that transforms simple prompts into 30-second video commercials using ElevenLabs, Google's Gemini, and VEO 2. By leveraging Node.js, Express, and React, Morrison created a system where users can input a basic product idea and receive a fully produced ad, complete with AI-generated visuals, voiceovers, and sound effects. The process involves an eight-step pipeline, allowing users to refine each component with control over text, video, and audio elements. A key challenge was enhancing vague ideas into concrete concepts using Gemini, which required careful prompt engineering to avoid generic outputs. Morrison emphasizes the importance of prompt design, user experience, and the integration of multiple AI systems for efficiency and creativity. The tool's flexibility allows for various applications, from marketing prototypes to sponsored content, highlighting how AI can augment rather than replace creative processes.

ElevenLabs has introduced a significant enhancement to its Conversational AI platform by integrating true text and voice multimodality, allowing AI agents to process both spoken and typed inputs simultaneously. This development aims to improve user interactions by offering more natural, flexible, and effective communication across various scenarios, addressing the limitations of voice-only interactions such as transcription inaccuracies and user difficulties with complex inputs. By enabling users to switch between voice and text inputs seamlessly, the multimodal approach enhances interaction accuracy, user experience, and task completion rates, while promoting a more natural conversational flow. The platform provides easy configuration and deployment options, including widget, SDK, and WebSocket support, and benefits from existing innovations like high-quality voices, advanced speech models, and global infrastructure. The company anticipates that this feature will significantly enhance the capabilities and user experience of Conversational AI.

VisionStory, an AI video creation platform, transforms text into professional-grade videos with integrated visuals, editing, and voiceovers, aiming to simplify content creation for storytellers, educators, and marketers. Utilizing a suite of over 200 voices in 32 languages from ElevenLabs, VisionStory allows creators to tailor voice tone and style to various projects, such as YouTube content and explainer videos. Initially employing a mix of in-house models and third-party tools, VisionStory transitioned fully to ElevenLabs' voice technology stack, which includes Text to Speech, voice cloning, and denoising, enhancing their development capabilities and user personalization. This integration has significantly contributed to VisionStory's paid signups, making voice a central element of their monetization strategy, and has led to updates based on user feedback for more authentic and regionally accurate voices. ElevenLabs' comprehensive voice solutions have been instrumental in VisionStory's growth, providing natural-sounding voices that adapt to various contexts, thereby supporting a diverse global creator community.

ElevenLabs successfully scaled its AI voice platform from 11 to over 5,000 voices, creating a global marketplace that has distributed over $5 million to contributors, partly by leveraging Stripe's financial infrastructure. The company focused on developing a diverse voice marketplace that includes real voice clones across various accents, languages, and dialects, which was facilitated by Stripe's flexible billing and payment solutions. This approach allowed ElevenLabs to manage complex licensing, rights, and international compliance issues efficiently, enabling voice actors to maintain control while easily managing payments and fraud protection. The partnership with Stripe has allowed ElevenLabs to remain lean and agile, enabling rapid growth and development, with a single part-time engineer managing all payment operations. As AI evolves rapidly, ElevenLabs continues to adapt swiftly, benefiting from their strategic partnership with Stripe to maintain a competitive edge.

ElevenLabs has introduced a new batch calling feature for its Conversational AI platform, designed to automate and scale outbound voice communications, addressing the challenges of manual outbound calling. This feature allows the simultaneous initiation of multiple outbound calls, facilitating tasks such as sending alerts, conducting surveys, or delivering personalized messages with greater speed and consistency. It supports integration with existing telephony setups via Twilio or SIP trunking and offers advantages like increased call capacity, personalized communication through dynamic variables, resource optimization by reducing manual dialing, and consistent message delivery using AI agents. Key functionalities include recipient list uploads, dynamic variable integration, AI agent selection, scheduling options, real-time monitoring, detailed reporting, and API access for automated workflow management. Users are guided to ensure phone number connectivity and refer to documentation for setup and operational guidance.

LeadTailor.AI is revolutionizing sales outreach by utilizing ElevenLabs' advanced AI voice technology to create highly personalized and dynamic prospecting videos, which significantly outperform traditional static cold emails. The integration of ElevenLabs' realistic AI voices into each video allows LeadTailor.AI to achieve a conversion rate over ten times higher than industry standards, with 40% of prospects engaged and 10-25% converted into meetings, compared to the typical conversion rate of under 1%. The platform has expanded from using two AI voices to eight and now includes a voice cloning feature, enabling customers to use their own voices in videos effortlessly. This innovative approach has led to rapid scaling, with 40 customers onboarded in just one month, and prospects frequently praise the lifelike quality of the voices, which has been a key factor in enhancing engagement and achieving impressive outreach results.

ElevenLabs has been verified as one of the first launch partners on n8n Cloud, allowing developers to integrate its AI voice capabilities directly into the n8n editor without needing additional setup or coding. n8n, an open-source automation platform, enables users to connect AI, apps, and services through a visual editor, making it accessible for both developers and non-technical users. This integration enhances automation by embedding powerful tools into the editor, facilitating the creation of workflows with just a few clicks. The ElevenLabs node supports a range of applications, including speech restoration for communication disorders, dynamic ads, in-game dialogue, and educational tools using synthetic speech. Getting started with ElevenLabs on n8n is straightforward, and the node can be installed quickly to become a native part of any workflow. This partnership marks the beginning of ongoing enhancements to the ElevenLabs node, aiming to offer more features and controls, while engaging with the developer community to shape the future of voice automation.

Anna Neely from ElevenLabs describes how they developed a robust framework for testing and improving conversational AI agents, using their documentation assistant, El, as a case study. Their process involves establishing reliable evaluation criteria to monitor agent performance, focusing on criteria such as valid interactions, user satisfaction, and the agent's ability to solve user queries without hallucinating information. Once areas for improvement are identified, the Conversation Simulation API is employed to test these improvements through both full and partial conversation simulations. This structured testing approach, integrated with their CI/CD pipeline via ElevenLabs’ open APIs, allows for automated testing of updates, ensuring rapid iteration and preventing regressions. This methodology has significantly enhanced El's capabilities and provides a scalable framework applicable to other conversational agents.

Melania Trump has released an audiobook version of her memoir, titled "Melania," utilizing ElevenLabs' AI-powered voice technology to make it accessible to a global audience through the ElevenReader app. This innovative audiobook, initially available in English with plans for additional languages such as Spanish, Portuguese, and Hindi, features narration by Melania Trump's official AI voice, offering listeners the experience of hearing her story in her voice. ElevenLabs, known for its AI audio research and technology, has become a popular choice for authors and publishers aiming to transform written works into audio formats. The ElevenReader app hosts a vast collection of audiobooks, including works by notable figures like Dr. Maya Angelou and Deepak Chopra, with AI narration in their unique voices. Additionally, the app provides users the ability to listen to various texts read aloud by iconic voices, further enhancing the immersive audio experience.

Allô is a mobile-first business phone system designed for small to medium-sized businesses, developed by the Mobile First Company with backing from Lightspeed Venture Partners. The app transforms mobile devices into comprehensive business communication hubs, offering features such as AI-driven call summaries, spam blocking, and smart routing. It utilizes ElevenLabs' Text to Speech and Conversational AI technologies to provide fast, natural-sounding, multilingual voice interactions, featuring voices like Mylene for French users and Tanya for English speakers. The integration of these technologies allows Allô to deliver low-latency and highly responsive voice experiences across various telephony functions, including pre-recorded messages and real-time AI agents. After switching from a previous vendor, Allô saw significant improvements in voice quality and latency, leading to better user engagement and fewer complaints. The rapid development cycle and support from the ElevenLabs Startup Grants program enabled Allô to quickly bring their product from concept to production, demonstrating the potential of integrating advanced AI voice technologies in modern communication systems.

Particle, an AI-powered news app, has significantly enhanced user retention by implementing an AI Voice feature in collaboration with ElevenLabs. This feature, "Listen to the News," offers users an audio version of their personalized newsfeeds, which allows them to engage with content during activities like commuting or exercising. By utilizing ElevenLabs' advanced voice library and API, Particle has recreated the in-app reading experience in audio form, resulting in users who engage with the audio content staying twice as long as before. The natural and human-like quality of the AI voices has been a key factor in this success, as noted by Sara Beykpour, Co-Founder & CEO of Particle.

Fuel iX, in collaboration with ElevenLabs, has introduced Agent Trainer, an AI-driven tool designed to enhance the training of contact center agents by significantly reducing onboarding time and improving skill acquisition. This innovative solution addresses traditional training challenges such as slow ramp-up times and high attrition rates by utilizing lifelike voice and chat simulations powered by ElevenLabs' Conversational AI. These simulations provide realistic, scalable practice scenarios that help agents develop confidence and competency in handling complex customer interactions. The platform supports multilingual scenarios and provides training teams with actionable insights through detailed performance reports. By offering immersive, natural voice interactions, Agent Trainer not only accelerates learning but also improves overall customer experience, agent retention, and operational efficiency, allowing human trainers to focus more on coaching and performance development.

Jamie, an AI assistant for meetings, experienced a significant improvement in transcription speed and quality by replacing their custom pipeline with ElevenLabs Scribe, which offers high accuracy in both transcription and speaker diarization. Previously, Jamie's team faced challenges in maintaining their custom pipeline, which combined open-source models for transcription and speaker diarization, requiring substantial engineering efforts. With Scribe's ability to handle complex audio environments, including overlapping speech and non-verbal audio events, the integration process was quick and required minimal customization. The switch resulted in a tripling of transcription speed, processing a one-hour meeting in just 30–45 seconds, and eliminated speaker error complaints. This change not only reduced engineering overhead but also enhanced user satisfaction and increased meeting recordings per user, demonstrating Scribe's effectiveness across multiple languages, including English, German, Spanish, and Dutch.

ElevenLabs developed SB1, an innovative soundboard powered by their text-to-sound effects AI audio model, allowing users to generate an unlimited variety of sounds on demand. Unlike traditional soundboards that rely on static MP3 libraries, SB1 enables users to type descriptions, such as "soft ambient forest sounds," to produce custom sound effects. The core technology, ElevenLabs SFX API, facilitates this by processing user prompts to generate multiple sound variations, which can then be streamed or downloaded. Built as a web app using React and Tailwind CSS, SB1 supports both preset and custom modes, offering features like looping and keyboard bindings for live sound manipulation. The platform is versatile, catering to diverse applications from podcasting to game development, and is designed for scalability and low latency. This tool exemplifies the potential of generative AI in audio production, encouraging users to explore creative sound design without the constraints of traditional sample libraries.

Synthesia, a text-to-video platform, utilizes ElevenLabs' AI-driven voice technology to streamline video creation for enterprise teams by transforming scripts into complete talking head videos with AI avatars and natural-sounding voiceovers. This collaboration eliminates production delays and studio time, offering fast and high-quality content creation that aligns with the tone and context of the intended message. Teams can easily create personalized onboarding videos, generate product updates, and localize content for global markets without reshoots or complicated subtitle workarounds. ElevenLabs' extensive library of voices supports multilingual capabilities, allowing for seamless integration of natural-sounding speech into various media tools, including video editing software and training platforms.

HeyGen's Avatar IV and ElevenLabs Voice Changer offer a dynamic workflow for creating studio-quality AI characters by animating still images and enhancing voiceovers. This toolset allows storytellers, educators, and content creators to transform images into lifelike, animated characters with realistic facial movements and high-quality voiceovers, without requiring a studio environment. By uploading images to HeyGen and using ElevenLabs to refine recorded voiceovers, creators can produce engaging content for various applications such as films, YouTube videos, and educational materials. This accessible technology democratizes high-fidelity character creation, enabling anyone to generate animated characters with natural speech, thereby revolutionizing content production and localization.

Impact Voice Lab, part of the ElevenLabs Impact Program, helps individuals who have lost their ability to speak due to conditions like ALS or mouth cancer by connecting them with volunteers who clean and prepare old audio recordings to recreate their voices using text-to-speech technology. Volunteers, who don't need to be experts, edit these recordings by removing noise and enhancing clarity to facilitate the creation of a Professional Voice Clone. This initiative helps maintain social connections and alleviate the isolation often experienced by individuals with speech loss, as a familiar voice can aid in preserving personal identity and relationships. Since its launch in August 2024, the program has collaborated with over 150 non-profits and public sector organizations across the U.S. and in 20 other countries, aiming to help one million people reclaim their voice by integrating AI audio into various social, cultural, and educational contexts.

ElevenLabs' AI voiceovers and sound effects offer a powerful way to enhance Google's Veo 2 photorealistic videos, creating immersive experiences by transforming silent sequences into captivating stories. Veo 2, available in the Gemini web app, facilitates the easy generation of eight-second clips, but lacks narrative consistency, making voiceovers a crucial unifying element. ElevenLabs allows users to craft dynamic AI voiceovers in various languages, offering control over tone, pacing, and emotion to fit the video's mood. The process involves planning and scripting the narration to align with the video's timing, followed by generating the voiceover using ElevenLabs' text-to-speech technology. Users can then sync the voiceover with their clips using editing software and enhance the auditory experience with AI-generated sound effects from ElevenLabs' text-to-sfx generator, which allows for the creation of custom audio elements. These sound effects, such as ambient noises or specific sound prompts, can be layered to add realism and depth, ensuring the final video is both engaging and lifelike.

Text-to-Speech (TTS) technology, once limited by its robotic-sounding voices, has advanced significantly due to AI, making it increasingly popular for tasks like listening to news or reading books, and enhancing accessibility for users with visual impairments or dyslexia. The global TTS market is growing rapidly, driven by the need for better accessibility and the convenience it offers for multitasking in today's busy world. While Android's built-in TTS feature provides basic functionality, it has limitations such as mechanical voice quality and limited customization. ElevenReader emerges as a superior alternative, offering more natural-sounding, customizable voices in 32 languages, and supports various content formats including PDFs and eBooks. It is particularly beneficial for content creators and offers free usage with the option for paid upgrades.

AI narrator voices have become a dominant trend on TikTok and Instagram in 2025, offering creators an efficient way to enhance their content. These AI-generated voices provide unmatched speed, consistency, and creative control, allowing creators to produce high-quality audio quickly, which is crucial in today's crowded social media landscape. The trend, which includes a variety of voices from lively and humorous to soothing and authoritative, has reshaped viewer expectations and driven engagement across platforms. Creators can easily customize AI voiceovers to match their content style and brand identity, saving significant production time and effort. Platforms like ElevenLabs offer extensive customization options and professional-quality voices, making them popular choices for content creators looking to captivate audiences without needing extensive voice-acting experience.

ElevenLabs provides an advanced AI-powered text-to-speech (TTS) tool that enhances video creation by generating realistic, human-like voiceovers, which can be seamlessly integrated with CapCut, a popular and user-friendly video editing app. CapCut is widely used by creators due to its ease of use and high-quality video editing capabilities, but it lacks a built-in text-to-speech feature, necessitating the use of third-party software like ElevenLabs for audio narration. The TTS tool from ElevenLabs offers a variety of expressive voices, multilingual support, and customizable features such as Voice Cloning and Isolation, allowing users to create engaging and professional-sounding voiceovers for their projects. By combining ElevenLabs TTS with CapCut, users can easily elevate the auditory quality of their content to match its visual appeal, resulting in videos that are both visually and aurally engaging.

ElevenLabs has introduced European data residency for its Enterprise customers, allowing organizations operating in Europe to comply with local data sovereignty requirements while utilizing advanced voice AI technology. This initiative reinforces ElevenLabs' existing commitments to data security, including SOC2 certification, GDPR compliance, optional Zero Retention Mode, end-to-end encryption, and HIPAA-compliant configurations. The European data residency not only enhances compliance but also reduces latency for European users through localized processing. The company, founded by Polish childhood friends Mati Staniszewski and Piotr Dabkowski, maintains strong European roots with offices in London and Warsaw, and has invested $11 million in Poland. ElevenLabs partners with major European companies like Deutsche Telekom and Bertelsmann and is dedicated to advancing voice AI technology with a focus on security, privacy, and compliance.

ElevenLabs has introduced a new feature in their Studio product, allowing users to generate sound effects (SFX) using a Text to SFX model integrated into their longform editor. This functionality provides users with the ability to describe a sound and instantly receive four different sound generations to choose from, enhancing the depth and realism of audiobooks and scripts. The sound effects can either block other audio elements until they finish or play simultaneously, with adjustable volume levels for a seamless blend with voiceovers. ElevenLabs encourages users to explore these capabilities and create immersive soundscapes, while also highlighting other AI-driven innovations, such as an AI sales development representative that qualifies leads and voice cloning technology demonstrated in multiple Indian languages.

Eddi Khaytman, a successful affiliate marketer and AI creator, shares his strategies for earning the first $100 through ElevenLabs' affiliate program, emphasizing the importance of choosing high-demand AI products, being a genuine brand advocate, and creating valuable, educational content. He highlights the need to use products personally to offer authentic insights and to integrate affiliate links seamlessly into content that solves problems and engages audiences. Eddi underscores the significance of staying updated with platform changes to maintain content relevance and credibility, as well as optimizing for discovery through effective SEO practices. He advises building trust with audiences by disclosing affiliate links, being patient, and understanding that each piece of content is an investment in long-term audience growth.

The Dalí Museum in St. Petersburg, Florida, has launched "Dial Dalí," an innovative AI-powered experience that allows people in the U.S. to engage in real-time conversations with a recreated voice of Salvador Dalí, using ElevenLabs' advanced voice cloning technology. Developed in collaboration with creative agency Goodby Silverstein & Partners, the project uses archival recordings and interviews to meticulously replicate Dalí’s unique voice, accent, and speech patterns, enabling dynamic AI interactions via a phone line managed by Twilio. This initiative, which coincides with Dalí’s 120th birthday, aims to offer a blend of surrealism, insight, and wit, allowing callers to inquire about various topics or simply wish Dalí a happy birthday. The experience reflects ElevenLabs' broader mission to preserve and extend human voices for educational, entertainment, and cultural heritage purposes, promoting the potential of AI in reimagining and rediscovering iconic voices from the past.

In Colombia, individuals with ALS, a condition that affects speech, are utilizing ElevenLabs' voice cloning technology to maintain personal connections as their natural voices deteriorate. Orlando Ruiz, the founder of the ALS MND Association of Colombia, who was diagnosed with ALS in 2023, has benefited from a Colombian Spanish voice clone created by ElevenLabs, allowing him to communicate in a way that feels authentic and familiar. This technology has been praised for reducing feelings of social isolation, which often accompany voice loss in ALS patients, and for preserving personal relationships by enabling individuals to feel heard as themselves. Speech and language pathologist Richard Cave highlights the potential of this technology in maintaining social networks, which are crucial for mental wellbeing, noting that over 40% of ALS patients suffer from clinical depression exacerbated by social disconnection. While preliminary experiences like Orlando's suggest significant positive impacts, further research is needed to evaluate the effectiveness of voice AI in sustaining social connections and improving overall wellbeing.