ElevenLabs Black Friday 2026: Is the $99 Creator Year Worth It?
Episode 2 of my weekly show on AI voices. The main story: ElevenLabs' Black Friday deal, and the simple rule I use to decide if a year is worth it. After that, the $200,000 contest, the new ElevenAI desktop app, a court in Warsaw that called an AI voice in an ad a deepfake, what's left of Hume, and DeepL's live voice translation. Every story with its date and sources.
Disclosure: the ElevenLabs link marked * is an affiliate link. I'm an independent ElevenLabs affiliate.
Chapters
Each timestamp jumps to that part of the transcript below. “Load video” then starts right there.
- 00:00 Should you lock in a year of ElevenLabs?
- 00:58 ElevenLabs Black Friday: $99 Creator year, my rule
- 02:28 The $200,000 contest and how to enter
- 03:14 ElevenAI desktop app and Scribe v3
- 04:00 Court: an AI voice in an ad was a deepfake
- 05:41 Hume after the shutdown, DeepL Aure
- 06:37 Quick hits: Elvis v. Suno, ZONOS3-Mini, ChatGPT
- 07:22 Three things to do and what's coming
This week in 60 seconds
- ElevenLabs' Black Friday deal runs from November 16 to December 1 and is for annual plans only. The Creator plan costs $99 for the first year. Normally it's $22 a month.
- My rule: $99 is four and a half months of Creator at the normal price. So the question is whether you'll still use ElevenLabs in five months.
- ElevenLabs' contest The Year of More has $200,000 in prizes. A free account is enough, and entries close on November 30.
- ElevenAI is a new desktop app for dictation and meeting notes, in early access on the Mac. Under it runs Scribe v3, which ElevenLabs says makes 27 percent fewer word errors than Scribe v2.
- On November 13, a court in Warsaw ruled on an AI voice in an ad and called the ad a deepfake of voice actor Jarosław Łukomski. The company has to take the ads down, apologize and pay. The ruling isn't final.
- Hume's text-to-speech and EVI have been off since November 13. Its developer docs now list two other products.
- DeepL released Aure, a voice model that translates you live in meetings. DeepL says the translated voice keeps sounding like you.
- Quick hits: the Elvis lawsuit against Suno got bigger, Zyphra released the open model ZONOS3-Mini, and the old ChatGPT voice modes go away on December 11.
ElevenLabs Black Friday: should you lock in a year?
▶ In the video at 00:58Quick note first: I'm an independent ElevenLabs affiliate, so the ElevenLabs link in this section is an affiliate link. It doesn't change your price, and it doesn't change what I tell you.
Deal from November 16 to December 1, 2026
ElevenLabs started its Black Friday deal on November 16, and it runs until December 1. It's for annual plans only. The Creator plan, the one with professional voice cloning, costs $99 for the first year. That's $8.25 a month. Normally Creator costs $22 a month on monthly billing and $18.33 a month on yearly billing. After the first year, the regular yearly price applies.
Pro annual plans are about half price too. On November 18, the pricing page showed Pro on yearly billing at $37.13 a month in the first year, then $82.50. ElevenLabs' blog post says new Pro subscribers get 55 percent off annual plans, and monthly Pro subscribers who switch to annual get 50 percent off.
ElevenLabs says the discount is for “eligible subscribers” and doesn't say who that is. So check what your own account shows before you plan with it.
My rule: four and a half months
$99 is four and a half months of Creator at the normal monthly price. Is that a good discount? It is. The real question is: will you still use ElevenLabs in five months?
I use ElevenLabs every week, for the tests on this channel and for client work, so for me that's a yes. But look at the rest of this week's news before you answer. Hume switched off its voice products on November 13, six weeks after the notice. HeyGen is giving away its fast voice model, Voice Turbo, through the API until the end of the month. And an open model came out that runs in your browser. This market moves fast, and a year is a long time.
Why it matters for creators: If you've paid monthly for the last three months and actually used your credits, the annual plan is the cheaper way to do what you're already doing. If you're only curious, stay monthly or stay on the free plan. A discount on something you don't use is not a saving.
Sources: ElevenLabs blog: “The Year of More: $200,000 for what you create next”, November 16, 2026 · ElevenLabs pricing page (yearly billing, read November 18, 2026)
The $200,000 contest: The Year of More
▶ In the video at 02:28November 16 to November 30, 2026, 11:59 p.m. Eastern
The second part of the announcement is a contest. ElevenLabs calls it The Year of More. There's $200,000 in prizes: fifteen winners get $10,000 each, and there's one grand prize of $50,000.
How to enter, according to the official rules:
- Make an original video, 30 seconds to five minutes long, edited from several shots, clips or images.
- Use at least one sound and one visual made in ElevenCreative. That's ElevenLabs' tool for images, videos and audio.
- Post it on two of these platforms: Instagram, TikTok, YouTube, X or LinkedIn, with the hashtags #morethanvoice and #ElevenCreativeCompetitionVol2.
- Send in the entry form within 24 hours of your second post.
A free account is enough, you don't need a paid plan. You have to be at least 18, and residents of a few places are excluded, Italy, Brazil and Quebec among them. ElevenLabs' blog post also asks for the competition watermark and for tagging @elevenlabs and @elevencreative. The blog post and the rules differ on some dates, so I go by the official rules. Read them before you post.
Why it matters for creators: If you already make short videos, this is one video, two posts and one form. And entering costs nothing.
Sources: ElevenLabs: official rules of The Year of More · ElevenLabs: contest page · ElevenLabs blog: “The Year of More: $200,000 for what you create next”, November 16, 2026
ElevenAI desktop app and Scribe v3
▶ In the video at 03:14Summit on November 11, 2026, changelog of November 9
The deal didn't come alone. On November 11, ElevenLabs held its summit in New York, and the main announcement wasn't a voice. It's a desktop app called ElevenAI. It's in early access, and right now it's for the Mac. You hold a key and dictate into any app, and it takes notes in your meetings.
ElevenLabs says it doesn't store your audio, that your transcript history stays on your Mac, and that the app shows a notice when it's taking notes. That's their description. I haven't tested it yet.
Under it runs a new speech-to-text model, Scribe v3. ElevenLabs says it makes 27 percent fewer word errors than Scribe v2, measured on the public benchmark FLEURS across 102 languages. That's their number.
Why it matters for creators: If you write scripts by talking, or you need subtitles, speech to text just got better. And when a tool takes notes in a call, tell people yourself. Don't rely on a small notice.
Sources: ElevenLabs: ElevenAI · ElevenLabs blog on the New York summit, November 11, 2026 · ElevenLabs changelog, November 9, 2026 (Scribe v3)
A court in Warsaw: an AI voice in an ad was a deepfake
▶ In the video at 04:00Last week's main story was the podcaster Chris Williamson and Archer, a voice in the ElevenLabs voice library that, he says, sounds like him (episode 1). Two things happened since.
ElevenLabs reports when clone samples match a protected library voice
Changelog of November 9, 2026
One line in ElevenLabs' changelog, their public list of updates: when a developer creates an instant voice clone, the system reports if the samples match a protected voice in the voice library. The changelog doesn't say what happens after a match. ElevenLabs doesn't connect this to Archer, and I won't either. But it's the same question from the other side.
Source: ElevenLabs changelog, November 9, 2026 (Voice Library matches)
The ruling
Ruling of November 13, 2026, Warsaw Regional Court, not final
On November 13, a court in Warsaw ruled on an AI voice in an ad. According to the reports, the judge called it the first dispute of its kind in Poland. JFC Polska, a company that makes home sewage treatment tanks, ran ads with an AI narrator. The voice actor Jarosław Łukomski recognized his own voice in them. He says he never recorded those ads.
A court expert called the two voices deceptively similar, and the judge called the ad a deepfake. The company was ordered to take the ads down, apologize, and pay 30,000 złoty in compensation plus a further amount of more than 17,000 złoty. Together that's about $12,000. The ruling isn't final. The company can appeal.
I haven't found a written judgment. Everything here comes from press reports, most of them based on the Polish Press Agency.
Where did the voice come from?
According to Polish media reports, a designer working for the company told the court he used ElevenLabs and changed a voice from its voice library. The court didn't establish how that library voice was made, the report says. That's reported testimony, not a court finding. ElevenLabs was not a party in this case, and I haven't found a statement from them on it (as of November 18, 2026).
It's not only Poland
A court in Berlin decided a similar case last year. On August 20, 2025, the Berlin Regional Court II ruled that a YouTuber had to pay the voice actor Manfred Lehmann, the German voice of Bruce Willis, a licence fee of 4,000 euros for an AI imitation of his voice (case 2 O 202/24).
Why it matters for creators: The company that published the ad was sued, not the tool. So a voice from a library doesn't protect you. If it sounds like a real person, that's your risk.
Sources: RMF24, citing the Polish Press Agency, on the ruling of November 13, 2026 (Polish) · dziennik.pl on the same ruling (Polish) · Bezprawnik on the designer's testimony, November 15, 2026 (Polish) · Voiceover Herald: English summary of the ruling · LTO on the Berlin ruling of August 20, 2025 (German)
Hume after the shutdown, and DeepL's live voice translation
▶ In the video at 05:41Hume: the voice products are gone
Off since November 13, 2026, developer docs read on November 18
Last week I said: export today. Since November 13, Hume's text-to-speech and EVI, its conversational voice API, are off. When I checked Hume's developer docs this week, only two products were listed: Audio Decisions and Expression Measurement. No Octave, no EVI. The company is still there. The voice products aren't. And Hume's notice said that data left in the account is permanently deleted.
Why it matters for creators: Hume gave six weeks' notice. So keep your source recordings and your scripts on your own drive, not just inside a tool.
Source: Hume developer docs index (read November 18, 2026)
DeepL Aure: live translation in your own voice
November 11, 2026
You probably know DeepL from text translation. On November 11, they released a voice model called Aure. It translates you live, in online meetings and in conversations in person. DeepL says the translated voice keeps sounding like you and follows your tone and your pace. It starts with eight languages: English, German, Spanish, French, Italian, Japanese, Portuguese and Korean. The press release names no price and no benchmark.
Why it matters for creators: That's not a dubbing tool for finished videos. It's for live calls, like an interview with a guest who speaks another language. I'm German, so English and German is the first pair I'd use.
Quick hits
▶ In the video at 06:37The Elvis lawsuit against Suno got bigger
Reported November 16, 2026
The lawsuit of Elvis Presley Enterprises against Suno, from last week's episode, got bigger. According to Digital Music News, an amended complaint lists at least 42 more Suno songs as examples. I couldn't read the new filing myself. It's still an accusation and not a ruling.
Sources: Digital Music News on the amended complaint, November 16, 2026 · Original complaint, Elvis Presley Enterprises et al. v. Suno (PDF, hosted by Music Business Worldwide)
ZONOS3-Mini: an open text-to-speech model that clones voices
Model card read on November 17, 2026
Zyphra released ZONOS3-Mini, an open text-to-speech model. The license is Apache 2.0, so commercial use is allowed. It covers 43 languages, it clones a voice from a clip of a few seconds, and Zyphra says it runs in the browser. Same rule as always: only clone a voice you have the rights to.
Source: Hugging Face: Zyphra ZONOS3-Mini model card
The old ChatGPT voice modes go away on December 11
Announced November 10, 2026
OpenAI is retiring the Advanced and Standard voice modes in ChatGPT on December 11. Everyone moves to the newer one, GPT-Live. If you use ChatGPT voice in your content, record your comparison before that.
Source: OpenAI Help Center: Retiring Advanced and Standard voice in ChatGPT
HeyGen Voice Turbo is free until November 30
HeyGen blog, updated November 2, 2026 · in the video at 00:58
HeyGen's fast voice model, Voice Turbo, is free through HeyGen's API until November 30.
Source: HeyGen blog: HeyGen Voice Turbo
Three things to do this week
▶ In the video at 07:22If you remember one thing from this week: a discount is only a saving if you would have paid anyway.
- Look at your usage from the last three months before you buy any annual plan.
- If you make videos anyway, enter the contest before November 30.
- If you publish with a library voice, listen to it again and ask who it sounds like.
Coming up
Dates worth knowing about, as of November 18, 2026. The current list is on the AI Voice Weekly page.
- Nov 27Black Friday itself, so expect more deals. My Black Friday overview for AI tools tracks what's announced.
- Nov 30ElevenLabs' contest The Year of More closes at 11:59 p.m. Eastern. On the same day, HeyGen Voice Turbo stops being free through HeyGen's API.
- Dec 1ElevenLabs' Black Friday deal ends: the annual Creator plan for $99 in the first year.
- Dec 11OpenAI retires the Advanced and Standard voice modes in ChatGPT. Everyone moves to GPT-Live.
Full transcript
From the English subtitles of the video, grouped by chapter. Timestamps open the video on YouTube.
00:00 Should you lock in a year of ElevenLabs?
On Monday, ElevenLabs started its Black Friday deal: the Creator plan for $99, for a whole year. And it put $200,000 on the table for people who make videos with its tools. So should you lock in a year of ElevenLabs right now? That's today's main story, and I'll give you the simple rule I use to decide. After that: ElevenLabs' new desktop app that takes notes in your meetings, what's left of Hume after the shutdown, and DeepL wants to translate you live, in your own voice. And last week I asked: who owns a voice that only sounds like you? A court in Warsaw just gave an answer, and it ordered the company to pay. I'm Lukas. I test AI voice tools on this channel, and every Thursday I go through what changed in AI voice and what it means if you create with it.
If that's useful to you, subscribe. Before we start: I'm an independent ElevenLabs affiliate. So the ElevenLabs link below is an affiliate link. It doesn't change your price, and it doesn't change what I tell you.
00:58 ElevenLabs Black Friday: $99 Creator year, my rule
Last week I said I'd cover the real Black Friday deals. Here's the first one. ElevenLabs started on November 16, and the deal runs until December 1. It's for annual plans only. The Creator plan, that's the one with the professional voice cloning, costs $99 for the year. Normally it's $22 a month. Pro annual plans are about half price too. The exact numbers are on screen right now. ElevenLabs says the discount is for eligible subscribers, and it doesn't say who that is. So check what your own account shows before you plan with it. Now, is it worth it? Here's my rule. $99 is four and a half months of Creator at the normal price.
Is this a good discount? It is. The question is: will you still use ElevenLabs in five months? I use ElevenLabs every week for this channel and for my customers, so for me that's a yes. But look at the rest of this week's news before you answer. Hume, the company behind the Octave voice model, switched off its voice products last Friday, six weeks after the notice. HeyGen is giving away its fast voice model, Voice Turbo, through the API until the end of the month. And last week, an open model came out that runs in your browser. This market moves fast, and a year is a long time. So if you've paid monthly for the last three months and you actually used your credits, the annual plan is the cheaper way to do what you're already doing.
If you're only curious, stay monthly, or stay on the free plan. A discount on something you don't use is not saving.
02:28 The $200,000 contest and how to enter
The second part is the contest. ElevenLabs calls it The Year of More. $200,000 in prizes: fifteen winners get $10,000 each, and there's one grand prize of $50,000. A free account is enough to enter. You make a video, 30 seconds to five minutes long, with at least one sound and one visual made in ElevenCreative. That's ElevenLabs' tool for images, videos and audio. Then you post it and send in a form. The exact steps are on screen, and the official rules are linked below. Entries close on November 30. You have to be 18, and a few countries are excluded, Italy among them. What this means for you: if you already make short videos, this is one video, two posts and one form.
And entering costs nothing.
03:14 ElevenAI desktop app and Scribe v3
That deal didn't come alone. On November 11, ElevenLabs held its summit in New York, and the main announcement wasn't a voice. It's a desktop app called ElevenAI. It's in early access, and right now it's for the Mac. You hold a key and dictate into any app, and it takes notes in your meetings. ElevenLabs says it doesn't store your audio, and that the app shows a notice when it's taking notes. That's their description. I haven't tested it yet. Under it runs a new speech-to-text model, Scribe v3. ElevenLabs says it makes 27% fewer word errors than the version before. That's their number, on a public benchmark. What this means for you: if you write scripts by talking, or you need subtitles, speech to text just got better.
And when a tool takes notes in a call, tell people yourself. Don't rely on a small notice.
04:00 Court: an AI voice in an ad was a deepfake
Now, the question from last week. The main story was the podcaster Chris Williamson and Archer, a voice in the ElevenLabs voice library that, he says, sounds like him. That episode is on screen now. Two things happened since. First, one line in ElevenLabs' changelog, their public list of updates. Since November 9th, when a developer creates an instant voice clone, the system reports if the samples match a protected voice in the voice library. ElevenLabs doesn't connect this to Archer, and I won't either. But it's the same question from the other side. Second, the court. On November 13, a court in Warsaw ruled on AI voice in an ad. The judge said it's the first case of its kind in Poland.
A company that makes home sewage tanks ran ads with an AI narrator. The voice actor Jarosław Łukomski recognized his own voice in them. He says he never recorded those ads. A court expert called the two voices deceptively similar, and the judge called the ad a deepfake. The company was ordered to take the ads down, apologize, and pay the equivalent of about $12,000. The ruling isn't final. The company can appeal. Where did the voice come from? According to Polish media reports, the company's designer told the court he used ElevenLabs and changed a voice from its voice library. The court didn't establish how that library voice was made, the report says. ElevenLabs was not a party in this case, and I haven't found a statement from them on this.
What this means for you: the company that published the ad was sued, not the tool. So a voice from a library doesn't protect you. If it sounds like a real person, that's your risk. And it's not only Poland. A court in Berlin decided a similar case last year: a YouTuber had to pay a voice actor for an AI imitation of his voice.
05:41 Hume after the shutdown, DeepL Aure
Now Hume. Last week I said: export today. Since November 13, Hume's text-to-speech and EVI, its conversational voice API, are off. When I checked Hume's developer docs this week, only two products were listed. No Octave, no EVI. The company is still there. The voice products aren't. And Hume's notice said that data left in the account is permanently deleted. So keep your source recordings and your scripts on your own drive. And DeepL. You probably know them from text translation. On November 11, they released a voice model called Aure. It translates you live, in meetings and conversations. DeepL says the translated voice keeps sounding like you, and follows your tone and your pace.
It starts with eight languages, English and German are among them. For creators, that's not a dubbing tool for finished videos. It's for live calls, like an interview with a guest who speaks another language. I'm German, so English and German is the first pair that I'd use.
06:37 Quick hits: Elvis v. Suno, ZONOS3-Mini, ChatGPT
A few quick ones. The Elvis lawsuit against Suno, from last week's episode, got bigger. According to Digital Music News, an amended complaint lists at least 42 more Suno songs as examples. It's still an accusation and not a ruling. Zyphra released ZONOS3-Mini, an open text-to-speech model. The license is Apache 2.0, so commercial use is allowed. It covers 43 languages, it clones a voice from a clip of a few seconds, and Zyphra says it runs in the browser. Same rule as always: only clone a voice you have the rights to. OpenAI is retiring the Advanced and Standard voice models in ChatGPT on December 11th. Everyone moves to the newer one, GPT-Live.
If you use ChatGPT voice in your content, record your comparison before that.
07:22 Three things to do and what's coming
So if you remember one thing from this week: a discount is only a saving if you would have paid anyway. Three things you can do this week. Look at your usage from the last three months before you buy any annual plan. If you make videos anyway, enter the contest before November 30. And if you publish with a library voice, listen to it again and ask who it sounds like. Coming up: November 27 is Black Friday itself, so expect way more deals. November 30th, the ElevenLabs contest closes and HeyGen Turbo stops being free. December 1st, the ElevenLabs deal ends, and December 11th, the old ChatGPT voices go away. All sources are down in the description.
And tell me in the comments: do you pay for your voice tool monthly or yearly? And bonus: which one is it? Thank you so much for watching. See you again next week.
Related
A new episode every Thursday
6 p.m. CET, 12 p.m. ET, on my YouTube channel. And tell me in the comments: do you pay for your voice tool monthly or yearly?
Subscribe on YouTube