The aspect of ElevenLabs that provides the most immense value to my workflow is the unparalleled natural realism and emotional depth of its AI voice generation. Using the Voice Design and Speech-to-Speech features regularly, I am consistently impressed by how the platform captures subtle human inflections, breathing paces, and precise pacing without the robotic artifacts common in other tools. This high-performance intelligence has completely transformed my production workflow; instead of spending days hiring voice talent or manually editing audio tracks, the intuitive UI allows me to generate and refine studio-quality narrations in minutes. Furthermore, the platform's robust API integrations make it incredibly seamless to deploy these voices directly into external applications, delivering a massive return on investment in both time saved and production quality. Review collected by and hosted on G2.com.

While the core voice generation is outstanding, the primary drawback lies in the performance consistency of long-form audio generation and the pricing model's impact on iterative workflows. When generating longer scripts, the emotional tone or pacing can occasionally drift dramatically mid-paragraph, forcing me to regenerate segments multiple times. Because ElevenLabs charges per-character for every single generation—including tweaks and minor corrections—this variance in performance can quickly exhaust monthly character quotas and diminish the overall ROI for larger projects. To solve this pain point, it would be a massive workflow improvement if the platform introduced a localized "preview" or "regenerate sentence" feature that doesn't charge the full token cost for minor structural adjustments, or added finer timeline controls within the UI to manually anchor shifting voice inflections before hitting generate. Review collected by and hosted on G2.com.

I really like ElevenLabs for its standout voice quality, which provides rich, emotional, and realistic voices. It's one of my favorite voice providers and it covers small country languages like Lithuanian, something most publicly available providers don't offer or only provide low-quality robotic voices. Review collected by and hosted on G2.com.

There are a couple of concerns I have with ElevenLabs. The STT part is still quite fresh. The newest models, like v3, are quite expensive and using it commercially is quite limited. V3 is the only model capable of working with the Lithuanian language, and it doesn't perform well in real-time voice applications like conversational agents. ElevenLabs is still not available at main cloud providers if you want to use it in close proximity and with a virtual private cloud setup. It was not trivial to start using it; generating an API key was not a direct funnel, and I struggled a bit to find out how to generate it on the webpage. The documentation should be served better. Review collected by and hosted on G2.com.

I love how ElevenLabs delivers realistic voices and quickly turns any text into speech. It helps me add emotion and accents in my daily work. The integration with other applications is smooth, and the performance stays fast, which makes it reliable for all kinds of audio projects. Review collected by and hosted on G2.com.

The pricing range for heavy use isn’t very budget-friendly, and the voice cloning often takes many tries to get it right. These are the main things I dislike right now, and I’d really like to see them improved so they can turn into positives. Review collected by and hosted on G2.com.

With ElevenLabs, I can tweak different voices to get the exact match I have in mind. Whether I want something for casual or professional use, it feels unique and flexible. I can also merge multiple voices together for voiceovers, which helps when I’m running tasks or turning an article into something I can listen to on the go. Review collected by and hosted on G2.com.

It can be a little inconsistent with pronunciation, especially for less common words or ones with unusual spellings. Sometimes it just ignores my changes or ends up messing up the sentence flow. Review collected by and hosted on G2.com.

What I like most about ElevenLabs is its natural-sounding voice quality and realistic speech synthesis. The generated voices sound remarkably human, with expressive intonation and clear pronunciation, making them suitable for a wide range of applications. Highly realistic AI voices with natural tone and emotion. Voice cloning capabilities for personalized voice generation. Support for multiple languages and accents. Low-latency text-to-speech suitable for real-time applications. Easy-to-use API and developer-friendly integration. For me, the most valuable feature is the voice quality. Whether I'm creating voiceovers, AI assistants, or customer interaction flows, the speech sounds natural enough to provide a much better user experience than traditional text-to-speech systems. The biggest benefit is that it enables the creation of professional-quality voice content quickly, reducing the need for manual voice recording while maintaining a natural and engaging listening experience. Review collected by and hosted on G2.com.

The biggest drawback is pronunciation consistency. While the voices sound very natural overall, uncommon names, abbreviations, or industry-specific terms sometimes need phonetic adjustments or prompt tweaks to sound exactly right. Fine-tuning tone, pacing, or pronunciation sometimes requires multiple iterations. Usage limits on lower-tier plans can be restrictive for frequent users. Certain words, names, or technical terms may occasionally be mispronounced without manual adjustments. Review collected by and hosted on G2.com.

I use ElevenLabs to create voiceovers for my German language learning videos. The text-to-speech function is very helpful because I can prepare educational content and let the AI voices speak the text clearly and naturally. I especially like the voice quality and the option to make the speech more expressive and emotional, because it helps my videos sound more interesting and less robotic.

The website is also easy to use. The interface is clear, and I was able to understand the main functions without a long onboarding process. Even without a premium subscription, ElevenLabs gives me useful tools for creating content. The audio generation works well for my needs, and I can use the results in my video workflow for platforms like YouTube, TikTok, and Instagram. Review collected by and hosted on G2.com.

What I would like to see improved is the availability and quality of some AI voices. There are not always enough good voices or bots available for the languages and accents I need. Some of the best voices seem to be connected to paid plans, which is understandable, but it can be limiting for users who are still testing the platform or creating content on a smaller scale.

I also think some voices could be improved in terms of volume, pronunciation, accent quality, and emotional expression in different languages. For language-learning content, this is especially important because the voice should sound clear, natural, and easy to understand.

I have also used the dubbing feature, and while it is useful, I feel that after some recent updates it became less convenient for me. In my experience, the dubbing sometimes requires more credits than before, and the final result does not always feel as good as it used to. I would like to see better consistency in dubbing quality and credit usage. Review collected by and hosted on G2.com.

ElevenLabs has been a very useful tool in our design and marketing workflow. I’ve been using it for about six months, mainly to generate voices for characters and to clone voices when we need to correct or adjust audio in videos.

In our company, we’ve used it specifically to create the voice of a mascot for a marketing campaign at Gas Natural del Norte, particularly for World Cup-related raffles and promotional content. It has also been very helpful for fixing or completing voice-over lines in videos that were already produced, saving us time and avoiding the need to re-record entire audio tracks.

The interface is very easy to use, the results are surprisingly realistic, and it allows us to produce high-quality voice content quickly without needing a full recording setup. I still feel there’s a lot more to explore in the platform, but so far it has been very impressive.---

If you want, I can make it sound more casual, more “corporate,” or optimize it specifically to pass G2 moderation even better. Review collected by and hosted on G2.com.

Since the platform offers a wide range of capabilities, it can take some time to fully explore and understand all the available features. I’m still discovering new ways to use it within our marketing and content workflows.

Additionally, depending on the project, it may require some experimentation with prompts and settings to achieve the exact tone or style needed, especially when working with branded content or character voice Review collected by and hosted on G2.com.

I like the ability that it has to create voices of varying degrees and the now module-like ability to apply the creative process to any form of media, such as Social, Entertainment, and Gaming. Review collected by and hosted on G2.com.

Not that it is a dislike, there are so many sections to explore, and a casual user can get lost in the plethora of utilities. Maybe a section that shows each area and how it can be used, probably using current users as virtual trainers. This is without having to leave the platform. Review collected by and hosted on G2.com.

Voice quality is genuinely hard to tell apart from a real person I use it for voiceovers and the output is usable in like 2 minutes without tweaking much Review collected by and hosted on G2.com.

The free tier is pretty limited on characters so you hit the cap fast if you're doing anything longer than a short clip Wish the pricing was a bit more flexible Review collected by and hosted on G2.com.

I like how simple and intuitive ElevenLabs is, especially the way it can be integrated with third-party applications. It is also very useful for the automatic generation of audio in my multimedia presentations, which greatly facilitates the work and significantly shortens the time. The initial setup was very simple. From selecting voices on the ElevenLabs page to programming with Laravel, everything was quite easy, using the guide as an assistant. I love that you can configure the language and voice parameters to be generated from the code. Review collected by and hosted on G2.com.

In the generation of audio for the summaries of scientific articles, we have noticed that in certain audios lasting more than 2 minutes, the tone and speed of the voice vary. It is not in all cases, I could say it may be in about 10% of the cases. Review collected by and hosted on G2.com.