Seedance 2.5 is here. Every model is included on every plan.
PricingLog inSign up
VoiceKokoro

Kokoro

Kokoro is a fast, open source speech model with twenty American voices and the lowest price per word on Dolly. Reach for it when you need a lot of speech quickly and cheaply.

Voices20 available
Make with Kokoro →View pricing

Added July 2026. Included on every plan.

Why this one

What makes Kokoro different

  1. The lowest cost per thousand charactersCheap enough to voice an entire library rather than a highlight.
  2. FastShort turnaround, which makes it the natural choice for a live loop with the agent.
  3. Twenty American voicesA compact set, split between female and male, easy to audition end to end.
  4. Open source lineageA well understood model with predictable behaviour.
  5. Natural for its priceIt reads noticeably better than its cost implies, particularly on short lines.
  6. Good for drafts that become finalsOften the placeholder is simply good enough to ship.

Use cases

Best for

  • Scratch voiceover
  • Bulk narration
  • Prototype audio
  • Social captions read aloud
  • Internal video
  • Agent iteration

Technique

Prompt tips

  1. Keep sentences shortIt handles compact sentences better than long subordinate clauses.
  2. Punctuate for breathCommas and periods are the pacing controls.
  3. Audition all twentyThe list is short enough to listen through once and then know your cast for good.
  4. Spell tricky words phoneticallyThe standard fix for a mispronunciation, and it works here.
  5. Use it for volumeAt this price, voice every variant rather than choosing one to voice.

Mechanism

How Kokoro works

You choose one of its voices and send the script; it renders speech from text and bills by characters sent. It is a smaller model than the commercial tiers, which is why it is fast and cheap and why it rewards short, plainly written sentences over long literary ones.

Getting started

How to use it on Dolly

  1. Pick KokoroChoose it in the voice studio, or just ask the agent for it by name.
  2. Write the promptDescribe what you want, in as much detail as the brief deserves.
  3. GenerateThe exact credit cost sits on the button before you press it, and a failed run refunds automatically.

Pricing

What it costs

Every generation shows its exact credit cost on the button before you run it.

per 1,000 characters
4 credits

One run is about 1/313 of a Starter month.

Failed generations refund automatically. See plans

Capabilities

Specs

Voices
20

Questions

Kokoro FAQ

How many voices does it have?
Twenty American voices, all listed on this page, split between female and male and short enough to audition end to end.
Is it good enough to publish?
Often, for internal video, social and prototypes. For brand narration, the ElevenLabs and Gemini tiers are a step up.
How is it priced?
Per thousand characters, and it is the cheapest speech model on Dolly.
Does it support other languages?
The voice list on this page is what is available here, and it is American English.
Why does a long sentence sound rushed?
Break it up. Short sentences and real punctuation fix pacing on this model.

More like this

Other voice models

ElevenLabs Multilingual V2ElevenLabs Multilingual V2 is the classic ElevenLabs voice model, with the largest voice library on Dolly and consistently lifelike delivery.ElevenLabs V3ElevenLabs V3 is the expressive ElevenLabs model, capable of multi voice dialogue in a single generation.Gemini 2.5 Pro TTSGemini 2.5 Pro TTS is Google's studio quality speech model, with a curated voice set and multi speaker scene support.

Ready to make with Kokoro?

Every model is included on every plan. The price sits on the button before you run it.

Start making →