TL;DR
ElevenLabs takes under 5 minutes to get your first voiceover. Sign up → choose a voice from the library → paste your script → generate. No credit card for the free plan. This guide covers everything beyond that: voice settings, cloning, and the Projects feature for long-form audio.
June 2026 update
ElevenLabs released two new generation models: Flash v2.5 (ultra-fast, low latency — best for real-time streaming) and Multilingual v3 (highest quality for non-English content). The voice settings interface is now consolidated under a single "Voice Settings" panel in the editor. IVC (Instant Voice Clone) is available on the Starter plan ($5/mo) and above — unchanged.
Part 1: Generate your first voiceover (5 minutes)
Create a free account
Go to ElevenLabs.io and sign up with email. No credit card required. You'll land on the main Speech Synthesis interface immediately. Free plan gives you 10,000 characters per month — about 8 minutes of audio.
Pick a voice from the library
Click the "Voice" dropdown. The default library has 10 voices on free (3,000+ on paid plans). Click "Preview" next to any voice to hear a sample before generating. For YouTube/content: "Adam," "Rachel," and "Domi" consistently rank well with audiences. Try a few and pick the one that fits your content tone.
Paste your script and adjust settings
Paste your text into the main box. Two sliders matter: Stability (0.5 is a good start — lower = more expressive, higher = more consistent) and Clarity + Similarity (keep at 75%). Leave Exaggeration at 0 unless you want dramatic delivery.
Generate and download
Hit "Generate." It usually takes 3-10 seconds for a paragraph. You'll hear the result in the browser player. If you want to regenerate with slightly different settings — just tweak and hit generate again. Each generation costs characters from your monthly quota. When you're happy, click the download icon to get the MP3.
Voice settings that actually matter
Most people leave these on defaults and wonder why the voice sounds slightly off. Here's what the sliders actually do:
Stability (0.0 – 1.0)
Lower = more expressive, emotional, variable. Higher = more consistent, robotic. For narration: 0.4–0.6. For dramatic/storytelling content: 0.2–0.4. Never go below 0.15 — it gets weird.
Clarity + Similarity Enhancement (0.0 – 1.0)
Higher = closer to the original voice sample, cleaner audio. Lower = might sound more natural but slightly different from the voice profile. Keep at 0.7–0.8 unless you're getting artifacts.
Style Exaggeration (0.0 – 1.0)
Amplifies the emotional style of the voice. 0 for normal narration. 0.3-0.5 for persuasive sales-style audio. Anything above 0.6 tends to sound overdone.
Part 2: Clone your own voice (Starter plan, $5/mo+)
Voice cloning is on Starter ($5/mo) and up. This is where ElevenLabs gets genuinely powerful — you record a few minutes of your own voice, and ElevenLabs makes an AI version you can type text into. Your audience hears you. You didn't record anything.
Record your voice sample
Record 1–3 minutes of yourself reading clearly. Use a decent microphone in a quiet room. The content doesn't matter — ElevenLabs recommends reading a book passage or article. Avoid music, background noise, or multiple speakers in the sample.
Go to Voices → Add a new voice → Instant Voice Clone
Upload your audio file (MP3 or WAV). Name it. Hit "Add Voice." ElevenLabs processes it in under a minute. You now have an AI clone of your voice in your library.
Test it with a few sentences
Type 2-3 sentences, pick your cloned voice, generate. If it sounds slightly off — pitch, accent — you can improve it by uploading more/better samples. More samples = better clone, generally up to 30 minutes of audio.
Use it for all your content going forward
Now you can produce video narration, podcast episodes, course voiceovers — without sitting in front of a mic every time. Write the script, generate, done.
Part 3: Long-form audio with Projects (Creator plan)
The standard Speech Synthesis interface is good for up to a few paragraphs at a time. If you're making a full YouTube video script, a podcast episode, or an audiobook chapter — use Projects instead.
Projects lets you import a full document (paste or upload), assign different voices to different speakers, and generate the whole thing in sections. You can regenerate individual sentences without redoing the whole thing. It keeps your character count usage much lower than doing it paragraph by paragraph in Speech Synthesis.
Projects is available on Creator ($22/mo) and above. If you're making any kind of long-form content regularly, it's worth it.
Tips that save characters and improve quality
Use punctuation to control pacing. A comma creates a short pause. A period creates a longer one. An em dash (—) creates a breath. Don't rely on the model to pace itself — punctuate the way you'd naturally speak.
Spell out numbers and abbreviations. "In 2026" reads fine. "$47.99/month" can cause weird cadence — write "forty-seven ninety-nine per month" if it sounds off.
Generate in sections, not in one massive block. 500–800 words per generation gives better results than dumping 3,000 words at once. The model maintains consistency better in shorter chunks.
Download generations you like immediately. Your generation history stays accessible but it's cleaner to download and store locally as you go.
Try It Yourself — Free
Sign up free, pick a voice, paste your first script. You'll have AI audio in 5 minutes.
Start Free on ElevenLabs →Affiliate link — I earn a commission at no cost to you.
More ElevenLabs guides
Frequently asked questions
How do I start using ElevenLabs?
Go to elevenlabs.io → create a free account → select a voice → paste your script → Generate. First voiceover in under 2 minutes. See the ElevenLabs review for a full feature breakdown.
How do I clone my voice?
Voices → Add a new voice → Instant Voice Clone. Upload 1 minute of clear audio. Clone is trained immediately. Available on Starter plan ($5/month) and above.
What is ElevenLabs Projects?
Long-form audio generation — paste a full script and generate it as one complete audio file. Avoids the chunk-limit of the standard editor. Available on Creator plan and above.
Why does my voice sound robotic?
Lower Stability to 40–50% and Similarity to 70–75% for more natural output. High stability values cause monotone delivery. Also try different voices from the library.
Can I use ElevenLabs voices commercially?
Yes, on paid plans (Starter $5/month and above). Free plan is personal use only. Voice cloning of other people requires their explicit consent.
Written by Shash
Founder, Infinfy Solutions. I use these tools on real work, then write about what actually happened.