cd /news/ai-products/can-elevenlabs-fix-my-cold · home › topics › ai-products › article
[ARTICLE · art-164467] src=workflows-of-ai.com ↗ pub= topic=ai-products verified=true sentiment=↑ positive

Can ElevenLabs fix my cold?

ElevenLabs released Eleven v4 and Eleven v4 Turbo on 28 September 2026, adding support for more than 90 languages (up from 70 in v3), stackable audio tags such as [excited, happy], instant voice clones from about 10 seconds of audio, and a v4 Turbo mode for voice agents with a median of about 150 ms to first speech, per ElevenLabs' own measurement. A first-person test by a writer who recorded 30 seconds of a strained, cold-stricken voice found that a Professional Voice Clone — which ElevenLabs trains on at least 30 minutes of clean speech and recommends 2-3 hours for — sounded more like the writer than the original recording, while the Voice Changer preserved the source speech's speed, cadence and pauses across four other voices.

by read10 min views1 publishedNov 5, 2026
Can ElevenLabs fix my cold?
Image: Workflows-Of-Ai (auto-discovered)

AI

I came back from Japan sick, recorded my strained voice and ran it through a voice clone, other voices, audio tags, music and dubbing. Here is every clip.

Turn your volume up before you start. This post is mostly audio clips.

I was in Japan on vacation for a couple of weeks, and as usual for me when I wind down, I got sick. My voice was very strained, so I recorded 30 seconds of it and spent a morning seeing how far I could push ElevenLabs v4 with it: my voice clone, other voices, audio tags, a movie trailer, a Swedish dub and one detour to HeyGen.

Every clip is below, in the order I made them.

The recording #

This is the starting point: 30 seconds of me, sick, with no clean-up. Every test after this one uses this file or the same words.

Real recording

My strained voice

My strained voice

0:30 Real recording

Press play to listen. Your browser starts sound only after you click or tap the page.

  1. So, I was in Japan for vacation for a couple of weeks, and as usual for me when I wind down, I end up getting sick.
  2. So, as you probably can hear, my voice is very strained.
  3. But let's see if we can have AI fix this for me.
  4. We will be using this audio file and some other inputs to see exactly how far we can push ElevenLabs version 4.

What's new in v4 #

ElevenLabs released Eleven v4 and Eleven v4 Turbo on 28 September 2026. What changed, according to the ElevenLabs announcement and TechCrunch:

  • More than 90 languages, up from 70 in v3.
  • Audio tags can be stacked, like [excited, happy] , and the model follows them in order.
  • Instant voice clones from about 10 seconds of audio.
  • v4 Turbo is for voice agents, with a median of about 150 ms to the first speech. That is ElevenLabs' own measurement.

My voice clone #

First, the same words in my Professional Voice Clone. A Professional Voice Clone is trained on your own recordings: ElevenLabs asks for at least 30 minutes of clean speech and recommends closer to two or three hours. You can only clone your own voice, and since a model is fine-tuned on it, it takes a while before the clone is ready and you get an email when it is.

Cloning doesn't use v4, it uses its own models.

AI-generated · ElevenLabs Professional Voice Clone

My voice clone

My voice clone

0:30 AI-generated · ElevenLabs Professional Voice Clone

Press play to listen. Your browser starts sound only after you click or tap the page.

This actually sounds more like me than the original clip, where it is harder to recognize my voice. So the experiment works.

Let's do more advanced things that ElevenLabs offers.

Four voices that are not mine #

Then the same words in four other voices. This is the Voice Changer, so it keeps my actual speech one to one: the speed, the cadence, the s and where I put the emphasis. You can still hear my Swenglish intonation in them. Jessica is one of the stock voices in ElevenLabs. For the kawaii one I turned similarity all the way down and the style up.

AI-generated · ElevenLabs

Jessica

Jessica

0:30 AI-generated · ElevenLabs

Press play to listen. Your browser starts sound only after you click or tap the page.

AI-generated · ElevenLabs

Deep voice

Deep voice

0:30 AI-generated · ElevenLabs

Press play to listen. Your browser starts sound only after you click or tap the page.

AI-generated · ElevenLabs

Posh English

Posh English

0:30 AI-generated · ElevenLabs

Press play to listen. Your browser starts sound only after you click or tap the page.

AI-generated · ElevenLabs

Kawaii, no similarity, high style

Kawaii, no similarity, high style

0:30 AI-generated · ElevenLabs

Press play to listen. Your browser starts sound only after you click or tap the page.

Story mode with audio tags #

Audio tags are the big thing in v4, so I took a neutral British voice and gave it the script three ways: plain, the same words with exaggerated tags, and a rewritten campfire story. Then I gave the story mode version to my own clone.

The tags change more than the tone. The plain read takes 21 seconds, and the same words with tags take 28.

AI-generated · ElevenLabs

Neutral British, plain

Neutral British, plain

0:21 AI-generated · ElevenLabs

Press play to listen. Your browser starts sound only after you click or tap the page.

AI-generated · ElevenLabs

Neutral British, story mode

Neutral British, story mode

0:28 AI-generated · ElevenLabs

Press play to listen. Your browser starts sound only after you click or tap the page.

  1. [overly dramatic, theatrical] So... I was in JAPAN.
  2. [gasps] For vacation! For a couple of weeks!
  3. [sighs heavily] And as usual for me... when I wind down...
  4. [voice cracking] I end up getting SICK.
  5. [despairing] So, as you probably can hear... my voice is very...
  6. [coughs] ...strained.
  7. [suddenly hopeful, excited] But let's see if we can have AI fix this for me!
  8. [building excitement] We will be using this audio file and some other inputs to see exactly how far we can push
  9. [shouting triumphantly] ElevenLabs version 4!

AI-generated · ElevenLabs

My voice clone, story mode

My voice clone, story mode

0:28 AI-generated · ElevenLabs

Press play to listen. Your browser starts sound only after you click or tap the page.

  1. [overly dramatic, theatrical] So... I was in JAPAN.
  2. [gasps] For vacation! For a couple of weeks!
  3. [sighs heavily] And as usual for me... when I wind down...
  4. [voice cracking] I end up getting SICK.
  5. [despairing] So, as you probably can hear... my voice is very...
  6. [coughs] ...strained.
  7. [suddenly hopeful, excited] But let's see if we can have AI fix this for me!
  8. [building excitement] We will be using this audio file and some other inputs to see exactly how far we can push
  9. [shouting triumphantly] ElevenLabs version 4!

AI-generated · ElevenLabs

Neutral British, campfire story

Neutral British, campfire story

0:46 AI-generated · ElevenLabs

Press play to listen. Your browser starts sound only after you click or tap the page.

  1. [dramatic storyteller voice, slow and mysterious] Gather 'round, friends... for I have a tale to tell.
  2. [long ] It began... in Japan.
  3. [wind howling] Two glorious weeks of vacation. Temples. Ramen.
4. [excited] So. Much. Ramen.
5. [lowers voice, ominous] But then...
  1. [short ] as it always does... the moment I finally relaxed...
  2. [thunder rumbling] the sickness came for me.
  3. [hoarse, theatrical gasp] Listen...
  4. [coughs] ...listen to what it has done to my voice!
  5. [sobbing dramatically] Strained! Broken! A shadow of its former self!
  6. [sudden determination, rising energy] But I will not go quietly!
  7. [heroic music swells] Armed with nothing but this humble audio file and a handful of inputs, I shall summon...
  8. [long ] [booming, epic voice] ELEVENLABS. VERSION. FOUR!
  9. [whispers] ...and we'll see just how far it can be pushed.

In a world… #

Then a movie trailer. I generated the narration and the music separately in ElevenLabs, and the agent took care of mixing them.

AI-generated · ElevenLabs voice and music

The trailer

The trailer

0:36 AI-generated · ElevenLabs voice and music

Press play to listen. Your browser starts sound only after you click or tap the page.

  1. [deep, gravelly movie trailer voice] In a world... where vacations end in sickness...
2. [long ] one man... went to Japan.
3. [short ] He came back...
  1. [coughs] ...changed. His voice... strained.
  2. [intense, building] Now, with a single audio file... and a few desperate inputs... he will discover the limits of
6. [booming] ElevenLabs. Version. Four.
7. [short ] [whispers] Coming this fall.

Or listen to the two parts on their own:

The narration on its own

0:25 AI-generated · ElevenLabs

The music on its own

1:00 AI-generated · ElevenLabs music

På svenska #

ElevenLabs dubbing translates audio and video into more than 90 languages and keeps each speaker's emotion, timing and tone. It clones the original speaker automatically, so the dubbed version speaks the new language in your own voice. Like cloning, dubbing doesn't use v4, it uses its own models.

I'm Swedish, so dubbing is the easiest test for me to judge. ElevenLabs dubbed my English recording into Swedish, and then I recorded myself saying the same thing in Swedish.

The dub keeps the length of the original, 30 seconds. My own Swedish take is 29.

AI-generated · ElevenLabs dubbing

ElevenLabs dub, Swedish

ElevenLabs dub, Swedish

0:30 AI-generated · ElevenLabs dubbing

Press play to listen. Your browser starts sound only after you click or tap the page.

Real recording

Me, in actual Swedish

Me, in actual Swedish

0:29 Real recording

Press play to listen. Your browser starts sound only after you click or tap the page.

Where the dub and I differ
English original ElevenLabs dub Me in Swedish
--- --- ---
as usual for me … I end up getting sick som vanligt … slutade jag med att jag blev sjuk som vanligt … slutar det med att jag blir sjuk
how far we can push hur långt vi kan pressa hur långt vi kan gå med
ElevenLabs version 4 ElevenLabs v4 ElevenLabs version 4

The dub says hur långt vi kan pressa. That is not really how you say "how far we can push" in Swedish. It is not wrong per se, but you would say hur långt vi kan gå, which translates to "how far we can walk".

My cloned voice also talks with more of a Stockholm dialect, while my natural dialect is a toned-down Gothenburg accent.

And funnily enough, it pronounces AI wrong.

Detour: a talking photo #

This one is HeyGen, which has a partnership with ElevenLabs. I gave it one photo of me from a conference talk and my Swedish recording.

AI-generated · HeyGen

A talking photo with my Swedish recording

It looks hilarious. It is obviously not there yet, and it took the office chair behind me and used it as a car: about 13 seconds in, a chair back drives from one shoulder to the other.

Eleven moods of Drupal #

The last test: one line per mood, one tag per line, all about Drupal.

AI-generated · ElevenLabs

Eleven moods of Drupal

Eleven moods of Drupal

0:30 AI-generated · ElevenLabs

Press play to listen. Your browser starts sound only after you click or tap the page.

  1. [deep, gravelly movie trailer voice] Drupal is the future.
  2. [angry] Drupal's security team did what they do best, another Wednesday update.
3. [happy] Drupal is awesome!
4. [bored] Drupal is very stable.
  1. [sharp exhale] Drupal is what we should be running.
6. [excited] Drupal is great with AI!
7. [sensual] Drupal is for everyone!
  1. [screaming over the rooftop] Everyone should use Drupal
  2. [course and sniffling] Drupal even works when I'm sick
  3. [panting] Drupal is a great framework
11. [singing] Drupal [] -
12. [singing] I'm lovin it!

ElevenLabs in Drupal #

ElevenLabs is what I use on this blog for the accessibility tracks, the read-out version of each post, and for the podcasts that get generated for each blog post.

If you want ElevenLabs in your own Drupal site, we have the ElevenLabs module for the integration.

So, can ElevenLabs fix my cold? #

Yes. My voice clone sounds more like me than the original clip, where it is hard to recognize my voice. It really does a great job with everything, except the talking photo. We do not talk about the talking photo.

#

Chapters 6 min read

Other ways to consume this

Listen to this article

16:17

Articles read out (podcast feed)

Narrated slides: Can ElevenLabs fix my cold? #

The post in three sentences #

I recorded my strained voice after getting sick on vacation and ran it through ElevenLabs v4: my voice clone, four other voices, audio tags, a movie trailer and a Swedish dub. Every clip is in the post, marked as AI-generated or a real recording. A HeyGen detour turned one photo of me into a talking video, with an office chair that drives around behind me.

source & further reading
workflows-of-ai.com — original article There's a module for that I vibe coded a point-and-click game to learn about the Drupal AI Ecosystem Eval of 19 ways to crop images using LLMs, Jev and image classification models
── more in #ai-products 4 stories · sorted by recency
── more on @elevenlabs 3 stories trending now
sponsored brought to you by zahid.host 4,200+ EU-deployed projects
reading about agents? ship yours in a single git push.

Run your AI side-project on zahid.host

EU-based hosting, git-push deploys, automatic HTTPS, no cold starts. Free tier with a custom domain — perfect for shipping the agent you just read about.

$git push zahid main
→ Live at https://your-agent.zahid.host ✓
Get free account → Pricing
from €0/mo · no card required
LIVE [news/can-elevenlabs-fix-m…] indexed:0 read:10min 2026-11-05 · —