How to convert text to speech with AI (realistic voices)

1. Go to https://elevenlabs.io
2. Click Sign Up (top right).
3. You can sign up with your Google account or email.
4. Once inside, you’ll see the main panel with the Speech Synthesis (Text to Speech) tool in the center.

ElevenLabs automatically assigns you the free plan with 10,000 credits per month. Your available credits appear in the left sidebar.

ElevenLabs has a library with hundreds of voices in multiple languages. Before converting your text, choose the voice that best fits your project.

1. In the Speech Synthesis panel, click the voice selector (where the active voice name is shown).
2. The Voice Library opens: you can filter by language, gender, age, and use case (narration, news, conversational).
3. Click Preview next to any voice to hear a sample.
4. When you find the voice you like, click Add to My Voices and then select it in the main panel.

Quality voices for English:

• Use the English filter to find native-sounding voices.
• The “Narration” and “Audiobook” labels usually give cleaner, more measured voices.

1. In the Speech Synthesis panel, type or paste the text you want to convert into the large text field.
2. Make sure the voice you chose in the previous step is selected.
3. Click Generate (the button with the play icon).
4. ElevenLabs processes the text in a few seconds and plays the audio automatically.
5. Listen to the result. If something doesn’t sound right, you can adjust the controls in the next step before downloading.

Best practices for the text:

• Use punctuation marks: commas and periods create natural pauses.
• Write numbers as words if you want control over how they’re pronounced (“twenty dollars” instead of “20 dollars”).
• For names or terms in another language, write them phonetically if needed.

ElevenLabs has controls to fine-tune how the voice sounds before downloading the final audio.

Voice controls (in the settings panel):

• Stability: how much the voice expression varies. Higher = more uniform and predictable; lower = more expressive and natural.
• Similarity: how closely the generated voice matches the original sample. Increasing this gives more consistency.
• Style: adjusts the exaggeration of the performance. For narrations, keep this value low.
• Speaker Boost: improves voice clarity. Turn it on if the voice sounds a bit flat.

To adjust speed, use the speed parameter in advanced settings (between 0.7 and 1.3 times normal speed).

For pauses within the text, add an extra comma or use the tag <break time="1s" /> (compatible with high-quality voices).

Once the audio sounds the way you want:

1. Click the download button (downward arrow icon) next to the generated audio.
2. ElevenLabs downloads the file in MP3 format.
3. You can import that MP3 directly into your video editor (CapCut, Premiere, DaVinci Resolve), to your podcast platform, or use it in presentations.

Common uses for generated audio:

• Narration for YouTube or Reels videos
• Voice-over for PowerPoint or Canva presentations
• Intros and outros for podcasts
• Audiobooks and online courses

Remember: with the free plan the audio is for personal use only. For monetized or commercial projects you need the Creator plan or higher. Always confirm current terms at https://elevenlabs.io/pricing.

If you want to combine this audio with AI-generated video, the tutorial How to create videos with AI shows you how to produce the video clips you can pair with these voice-overs.

If you already have an account and want to generate your first audio right away, follow this quick path:

1. Go to https://elevenlabs.io and sign in.

2. Go to Speech Synthesis.

3. In the voice selector, choose any voice with the “English” filter and listen to the sample.

4. Paste this test text:

   “Artificial intelligence doesn’t replace you: it gives you superpowers. What used to take hours now takes minutes. This is your first step to building more with less effort.”

5. Click Generate and download the MP3.

The free plan requires no credit card. You have enough credits to generate several minutes of audio per month.

To add this narration to a video, learn to produce visual clips in the tutorial How to create videos with AI.

Problem	What to do
The voice mispronounces a word	Write it phonetically or use the SSML <phoneme> tag
The audio sounds monotone or robotic	Lower the Stability parameter and raise Style slightly
The voice makes a weird pause mid-sentence	Check your punctuation marks; an extra period or comma creates an unnecessary pause
I ran out of monthly credits	Free plan credits renew on your subscription anniversary date; consider the Starter plan if you need more
I can’t download the audio	Check that you’re signed in; some browsers block automatic downloads
The audio has background noise or cuts	Try a different voice or increase the Speaker Boost parameter