This post contains affiliate links. If you purchase through them, we may earn a commission at no extra cost to you. Read our How we review page to see how we check prices and features.
ElevenLabs text to speech speed is controlled by the speed setting: 1.0 is the default, values below 1.0 slow the voice down, and values above 1.0 speed it up. The official API reference confirms this for the voice_settings speed parameter. Speed changes pacing, not pitch. Plans start at $0/mo for Free (checked 4 Nov 2026).
Best for: anyone who needs slower narration for learning content or faster delivery for short-form video and ads.
Not for: anyone who wants pitch changes. Speed adjusts pacing only; it does not make a voice higher or lower.
Starting price: $0/mo for the Free plan with 10,000 credits per month (checked 4 Nov 2026).
Where the speed control lives
Speed lives in two places. In the ElevenLabs app, the Text to Speech interface has a speed slider among the voice settings, next to stability, similarity, and style exaggeration. Through the API, speed is the speed field inside voice_settings on the text-to-speech convert endpoint. Both do the same thing; the API just lets you set it per request instead of per voice.
Our ElevenLabs text to speech settings guide covers the full settings panel, including how speed interacts with the other three sliders. This post goes deep on speed alone: what the numbers mean, how to set them, and what they cost.
How the speed numbers work
The official API reference is precise about the semantics: speed defaults to 1.0, values less than 1.0 slow down the speech, and values greater than 1.0 speed it up. In the app, the slider runs from 0.7 to 1.2, with 1.0 as normal. A setting of 0.8 plays at roughly 80 percent of normal pace; 1.2 plays at roughly 120 percent.
The important nuance is what speed does not do. It changes pacing, not pitch. A voice slowed to 0.8 does not become deeper, and a voice sped to 1.2 does not become chipmunk-like. ElevenLabs time-stretches the delivery while keeping the voice’s character intact, which is why speed is safe to adjust on a brand voice without changing how the brand sounds.
How to change the speed, step by step
- Open Text to Speech (0 credits). Go to the ElevenLabs app and open the Text to Speech page. Pick the voice you want to adjust; speed is set per voice in the interface.
- Find the speed slider (0 credits). Open the voice settings. The speed slider sits with stability, similarity, and style exaggeration. The default position is 1.0.
- Drag to your target (0 credits). Move toward 0.7 for slower, more deliberate narration, or toward 1.2 for brisk delivery. Small moves matter: 0.9 versus 1.0 is clearly audible over a full minute.
- Preview a full paragraph (credits for the preview). Generate a test with a representative paragraph, not a single sentence. Pacing problems hide in short samples and show up at length. One credit is charged per character generated.
- Lock it per voice or per request (0 credits). In the app, the setting sticks to the voice until you change it. Through the API, pass speed inside voice_settings on each request to vary pacing without touching the saved voice.
- Generate the final audio (1 credit per character). Speed itself costs nothing extra; you pay the normal per-character rate. A 1,000-character script costs 1,000 credits at any speed.
Speed vs stability vs style: the settings work together
Speed is one of four voice settings, and they interact. Stability controls how consistent the delivery is; low stability at high speed can sound rushed and erratic, while high stability at low speed can sound flat. Similarity keeps the voice close to the original; style exaggeration pushes expressiveness. If a sped-up voice sounds unnatural, raise stability slightly before touching speed again.
The practical rule: set speed last. Dial in the voice, stability, and style first at 1.0, then adjust pacing. Changing speed first and the voice second means re-tuning pacing every time the voice changes, because different voices carry speed differently.
What speed costs (spoiler: nothing extra)
Speed is a setting, not a feature tier. It costs zero additional credits on every plan, including Free. You pay only the standard generation rate of one credit per character. The Free plan’s 10,000 monthly credits therefore produce the same amount of audio at 0.7 as at 1.2; the difference is the runtime of the audio, not the credit cost.
For reference, the current plan lineup is Free at $0/mo (10,000 credits/mo), Starter at $6/mo (30,000 credits), Creator at $22/mo (121,000 credits), and Pro at $99/mo (600,000 credits), all checked 4 Nov 2026. A limited-time $1 first-month Starter offer was shown until Nov 8 when checked on 4 Nov 2026. None of the plans charge extra for speed adjustments.
When to slow down and when to speed up
Slow down (0.8 to 0.9) for learning content, meditation, audiobooks for young listeners, and any script with dense terminology. Listeners need time to process, and a slightly slower pace raises comprehension without sounding broken. Speed up (1.1 to 1.2) for short-form video hooks, ad reads with tight time slots, and recap content where energy matters more than savoring each word.
Stay at 1.0 for conversational content, dialogue, and anything where the voice must sound like a person talking normally. Most listeners never notice 1.0; they notice deviations. Use speed as a deliberate choice for a format, not as a default tweak on every project.
Speed presets worth saving
Most creators end up reusing three speed settings. Save them as a personal preset list: 0.85 for explainer narration, 1.0 for conversational and dialogue content, and 1.15 for short-form hooks and ad reads. Write these in your project template, because the app stores speed per voice and you will otherwise re-tune it from memory every time. If you work through the API, define the three values as constants in your integration and pass the right one per content type. Consistent presets also make testing meaningful: when every explainer uses 0.85, you can compare voices and scripts without pacing as a confounding variable.
Pros
- Free on every plan, including the $0 Free tier; pacing costs nothing extra.
- Adjustable per voice in the app and per request through the API.
- Changes pacing without changing pitch, so brand voices stay recognizable.
- Small adjustments (0.9 or 1.1) are clearly audible without sounding artificial.
Cons
- Cannot change pitch; a separate concern needs a different voice, not speed.
- Extreme settings can sound unnatural, especially combined with low stability.
- The app slider caps at 0.7 to 1.2, so dramatic slow-motion or chipmunk effects are out.
- Speed is per voice in the app, which means configuring it separately for each voice you use.
Who should NOT use it
Skip speed adjustments if your goal is a different-sounding voice: speed will not make a voice deeper, brighter, younger, or older. Pick a different voice instead. Also skip it for sung or musical content, where timing is part of the performance; time-stretching a vocal line changes its relationship to the beat in ways a slider cannot fix.
Frequently asked questions
What is the default ElevenLabs text to speech speed?
1.0 is the default. The official API reference states that speed defaults to 1.0, values less than 1.0 slow down the speech, and values greater than 1.0 speed it up. In the app, the slider runs from 0.7 to 1.2.
Does changing the speed change the pitch?
No. Speed changes pacing, not pitch. A voice slowed to 0.8 keeps its character and does not become deeper; a voice sped to 1.2 does not become higher. If you want a different pitch, choose a different voice.
Does speed cost extra credits?
No. Speed is a free setting on every plan, including the Free tier. You pay only the standard one credit per character for the generated audio, regardless of the speed setting.
Can I set speed through the API?
Yes. Pass the speed field inside voice_settings on the text-to-speech convert endpoint. This sets pacing per request, which is useful when one integration needs different speeds for different content without changing saved voices.
Why does my sped-up voice sound unnatural?
Extreme speed combined with low stability is the usual cause. Raise stability slightly, pull speed back toward 1.0, and preview a full paragraph. Small adjustments like 1.1 usually sound natural; 1.2 is the edge of the comfortable range.
What speed should I use for audiobooks?
Stay near 1.0 for most audiobooks, or drop to 0.9 for dense nonfiction and young listeners. Narration slightly slower than conversational pace aids comprehension over long sessions without sounding artificially dragged.
Sources checked
- ElevenLabs official API reference, text-to-speech convert (voice_settings speed parameter: default 1.0, below slows, above speeds up) – https://elevenlabs.io/docs/api-reference/text-to-speech/convert – checked 4 Nov 2026
- ElevenLabs official text-to-speech page (plan credits, $1 Starter offer timing) – https://elevenlabs.io/text-to-speech – checked 4 Nov 2026
Last updated 4 Nov 2026. Prices checked on the official pricing page on that date; confirm at checkout.