First of all, thank you for the amazing work you’re doing. The platform has become an important part of my workflow, and I really appreciate the quality you’re pushing toward. Review collected by and hosted on G2.com.

1) Voice consistency across models

When I train a voice in one model and then try to use it with a younger model, the voice changes quite noticeably and no longer sounds like the person I trained. Maintaining the identity of the original voice is very important for my projects, so this shift makes it difficult to use different models.

2) Pronunciation of numbers / “3D-style” words

The trained voice often struggles with numbers or terms like “3D.” The delivery feels unnatural and sometimes includes strange pauses in the middle of the phrase.

3) Emotional variation on first generation

When I generate a text for the first time, the read can sound flat. But if I regenerate the exact same text again, it suddenly becomes more fluent, expressive, and dynamic. I would really love to get that level of quality on the first try, without needing multiple regenerations.

4) Very important – incomplete dubbing / untranslated parts

During dubbing, some words or short segments are occasionally left in the original language or not dubbed at all. This is critical for me, because I then have to cut those parts out in my editing project and re-dub them separately, which slows everything down a lot.

I hope this feedback is helpful. I truly enjoy using ElevenLabs and would love to see improvements in these areas, as they would make a huge difference in my daily work.

Thank you for your time and support. Review collected by and hosted on G2.com.

I use ElevenLabs for voice changing and generating voices for organic ads. It’s really fast and the API works smoothly. I like that it's started by Polish guys, which makes me proud of its success in the AI and SaaS market as a fellow Pole. Dubbing, voice changing, and especially generating voices with the new model V3 are great. The V3 model lets me clarify exactly what I want, and it adds a lot of value because the voices sound natural. Working with Heygen using the new model is also definitely effective. The initial setup was very easy, as voice cloning or creating is simple, which is a big advantage. Overall, I haven't found another product that works as well as ElevenLabs. Review collected by and hosted on G2.com.

I find that generating long scripts is a bit problematic because the voice tends to slow down when the script is a little longer. The voice changer could also use some improvement, especially in English and for multilingual support in other languages. Review collected by and hosted on G2.com.

I primarily use ElevenLabs for Text-to-Speech to create high-quality voiceover for my YouTube channel. I like that it doesn't just read the text; it actually interprets the context, which makes my YouTube narrations feel much more engaging and authentic. It inserts subtle breaths and variations in rhythm, enhancing the natural feel of my recordings. Additionally, the setup was very easy, taking just minutes. Review collected by and hosted on G2.com.

I sometimes notice the voice profile shifts slightly during a long narration. Review collected by and hosted on G2.com.

I absolutely love ElevenLabs for its exceptional voice agent capabilities, which enable seamless conversations with users on my platform. The voice agent's remarkable smoothness and low latency make the experience delightful, often leading users to forget they are interacting with an AI rather than a human. I also appreciate the incredible ease of integrating ElevenLabs into a React app through its SDK. Furthermore, the dashboard is impressive, allowing me to edit the agent easily and providing visual confirmations for each action. The conversation history dashboard and analytics are invaluable, offering insights and data that greatly assist in managing and optimizing the user interactions. Review collected by and hosted on G2.com.

I think the price could be lowered. I can devote more time towards customizing the agent and get it working for my scenario, which is anyway a must for applications to grow at scale. But at this price, customers would be expecting an out-of-the-box solution to work very well. So I believe ElevenLabs could reduce the price and actually get to higher scale to have similar or better revenue. Review collected by and hosted on G2.com.

One of the things I really like about ElevenLabs — especially the newer updates — is how much focus they’ve put on voice customization and expressiveness. They’ve added features that let you tailor the emotion, pacing, and personality of the voices more precisely than before, which makes the generated speech sound much more natural and engaging.

I also appreciate the improvements in multilingual support — not just basic pronunciation, but voices that handle intonation and rhythm in languages like Spanish, French, and Hindi much better than earlier versions. This makes it far more useful for global audiences and multilingual content creators.

Another new thing I like is the enhanced API flexibility, which allows developers to integrate ElevenLabs into a wider range of applications — from podcasts and audiobooks to real-time conversational agents — without losing quality or responsiveness. Review collected by and hosted on G2.com.

1. Better Multilingual Support

A lot of the core voices are optimized for English, but in some other languages or accents, the quality and naturalness can vary. Expanding more high-quality voices for global languages would be a big plus.

2. More Intuitive Editing Tools

The platform is very capable, but for non-technical users, some features like fine-tuning voice style, emotion, pacing, and inflection can feel a bit complex. More visual or guided tools could make editing easier.

3. Improved Context Awareness

In long passages of text, generated speech can occasionally sound slightly repetitive or neutral in tone. Increasing semantic expressiveness—especially with nuanced emotion transitions—would make outputs feel even more human. Review collected by and hosted on G2.com.

I use ElevenLabs primarily for text-to-speech and voice generation tasks, and I’ve found it to be a powerful and flexible voice AI platform. The voices sound smooth, natural, and expressive compared to many other TTS tools. One standout feature is the ability to train or clone custom voices, which saves time and makes content feel more professional. I appreciate the multilingual voice support as it offers several languages and accents, which is very useful for reaching a global audience. I like the exceptional voice quality and realism, which makes a huge difference compared to traditional text-to-speech tools. The voice cloning and customization feature allows me to maintain a strong, recognizable audio identity. I'm impressed by the emotional control and intonation that capture pauses, emphasis, and tone very well. I also find the fast generation speed and easy-to-use interface valuable, making it suitable for professional content. ElevenLabs stands out because it combines top-tier audio quality with practical features, allowing me to scale voice content without sacrificing realism. The initial setup was very straightforward with an intuitive platform, making the onboarding process smooth and beginner-friendly. Review collected by and hosted on G2.com.

While ElevenLabs is a powerful tool, there are a few areas where it could be improved. One of the main challenges is pricing at higher usage levels. As usage scales, the cost can increase quickly, which may be difficult for individual creators or small teams managing tight budgets. Another area is fine-grained control over pronunciation and pacing. Although overall voice quality is excellent, there are occasional mispronunciations or pacing issues that require manual text adjustments or trial-and-error to correct. I’ve also noticed that advanced features have a learning curve. Custom voice creation and detailed tuning can take time to fully understand, especially for new users who want precise control. Review collected by and hosted on G2.com.

I really like the choice of models ElevenLabs offers; it really stands out to me. The voice cloning capability and the RAG retrieval capability are also impressive, as is the ability to create multiple agents and voices. I think the breadth of the product is amazing. Additionally, the initial setup was very easy, and the voice-guided agent on the site is really intelligent. It's able to respond extremely well to specific questions and guide any setup-related requirements. Review collected by and hosted on G2.com.

The pricing is really, really expensive. To actually run products at the kind of pricing that they have. Review collected by and hosted on G2.com.

I use ElevenLabs to communicate with others after losing my ability to speak due to a brain injury. I appreciate that some relay apps work with the cloned voice ElevenLabs created for me, and this helps a lot compared to the very robotic-sounding voices of other text-to-speech software. The initial setup of ElevenLabs was easy. Review collected by and hosted on G2.com.

I'm still trying to figure out how to use ElevenLabs, and I haven't been able to return to work yet because I can't get my cloned voice to work on Zoom calls or Teams video calls. My cloned voice tone is off, and it doesn't pronounce numbers correctly. In some cases, like with the number 5, it doesn't pronounce them at all. Also, I'm unable to use it online or on phone relay programs like Nagish. Review collected by and hosted on G2.com.

I really like the voice quality of ElevenLabs. It's the best we've figured out, offering far better quality compared to other models we've tried. It handles interruptions very nicely with its turn-taking capabilities, something we've always wanted in our LLMs. I find the expressive voice features of ElevenLabs to be very nice. Review collected by and hosted on G2.com.

I think the pricing could be improved. While the quality is really nice, the cost is a bit too much for integrating all our workflows into ElevenLabs. If the pricing was better, it would help us migrate everything smoothly. Another thing is the lack of end-to-end logging and tracing. It would be helpful to have a way to distinguish any latency between our system and ElevenLabs or in generating the voice itself. These improvements would be nice, especially when scaling production systems. Review collected by and hosted on G2.com.

I really like the audio quality in ElevenLabs; it's top-notch. The ability to generate full songs with or without lyrics is incredible. It saves me from having to generate content by hand, which would involve hiring musicians or artists—this would be expensive and tedious. Plus, it fits seamlessly into the game development workflow. Review collected by and hosted on G2.com.

Occasionally, the instructions become lyrics in the content, which is a bug. I've heard some tracks sing 'create a 15-second clip...' which is amusing but not something that works for my product. Review collected by and hosted on G2.com.