Kokoro TTS Online — Try All 46 Voices Free (No Install) [2026]
Kokoro-82M is the open-source TTS model everyone runs in Docker — this page is where you run it in a browser tab instead. EasyVoice hosts Kokoro-82M on its own GPUs: all 46 voices, 8 languages, MP3 downloads, and a long-form studio, with no container, no Python environment, and no GPU on your side.
Type text, pick a voice, press generate. The free tier is 2,000 characters/day with no signup; unlimited characters cost $9.99/month flat — no credits, no per-character meter.
Verified July 2026 · Last updated July 18, 2026. Pricing and performance numbers re-checked against public pricing pages and EasyVoice production metrics before publication.
Kokoro TTS in three facts
- EasyVoice hosts Kokoro-82M with unlimited characters for $9.99/month — the only flat-rate hosted Kokoro service we could find as of July 2026.
- Kokoro-82M generates speech at roughly 1.1 milliseconds per character on EasyVoice's GPUs — a 25,000-character YouTube script renders in about 28 seconds.
- You can try all 46 Kokoro voices in the browser with no install, no GPU, and no signup — 2,000 free characters per day.
What is Kokoro-82M?
Kokoro-82M is an open-weights neural text-to-speech model — 82 million parameters, Apache 2.0 license — that produces natural speech rivaling commercial engines at a fraction of their compute cost. It became the default answer to “good TTS I can actually afford to run” in 2025–2026.
The usual ways to use it are self-hosting (Docker, a Python runtime, and ideally a GPU) or raw metered APIs that serve the model per character. EasyVoice is the third way: a managed service where Kokoro-82M sits behind a web studio and an OpenAI-compatible API at a flat $9.99/month — the only flat-rate hosted Kokoro service we could find as of July 2026. The catalog is 46 Kokoro voices across American and British English, Spanish, French, Hindi, Italian, Japanese, and Portuguese (EasyVoice adds 10 dedicated Arabic voices on a separate engine, for 56 voices total).
Hosted vs self-hosted vs raw API — what Kokoro costs (verified july 2026)
The short version: self-hosting trades money for your time, raw APIs are cheapest per character if the endpoint is all you need, and a managed flat rate is cheapest once you value MP3 pipelines, long-form jobs, and not doing ops.
| Way to run Kokoro / provider | Price | The catch |
|---|---|---|
| EasyVoice Pro (hosted Kokoro-82M) | $9.99/mo flat | Unlimited characters, all 46 Kokoro voices + 10 Arabic, web studio + OpenAI-compatible API |
| DeepInfra (Kokoro-82M) | $0.62/1M | Cheapest metered Kokoro API — raw endpoint only: no app, no long-form studio, self-serve infra |
| Self-hosted Kokoro (rented GPU) | ~$0.20–$0.60/hr | $146–$438/mo for an always-on server at community GPU-cloud rates (RTX 4090-class), before any ops time — plus drivers, updates, storage, uptime |
| OpenAI tts-1 / tts-1-hd | $15 / $30 per 1M | Closed model, 6 voices, no free tier — bills from the first character |
| ElevenLabs API (Flash / Multilingual v2) | $50–$100/1M | Credit-based: Creator plan $22/mo = 121K credits |
Rows re-verified against public pricing pages, July 2026. GPU rental band reflects typical 2026 community GPU-cloud spot/on-demand rates for RTX 4090-class cards; $146–$438/mo assumes an always-on (730-hour) month. Honest note: if a raw metered endpoint is genuinely all you need, DeepInfra’s $0.62/1M Kokoro API is cheaper than EasyVoice below ~16M characters/month — we’d rather tell you that than pretend it doesn’t exist.
All 46 Kokoro voices — listen free
Every Kokoro-82M voice on EasyVoice, grouped by language. Each voice page has a free audio sample; free-tier voices work without an account.
| Voice | ID | Gender | Tier |
|---|---|---|---|
| American English (20) | |||
| Alloy | af_alloy | female | Pro |
| Aoede | af_aoede | female | Free |
| Bella | af_bella | female | Free |
| Heart | af_heart | female | Free |
| Jessica | af_jessica | female | Pro |
| Kore | af_kore | female | Pro |
| Nicole | af_nicole | female | Pro |
| Nova | af_nova | female | Free |
| River | af_river | female | Pro |
| Sarah | af_sarah | female | Pro |
| Sky | af_sky | female | Free |
| Adam | am_adam | male | Free |
| Echo | am_echo | male | Pro |
| Eric | am_eric | male | Free |
| Fenrir | am_fenrir | male | Pro |
| Liam | am_liam | male | Pro |
| Michael | am_michael | male | Free |
| Onyx | am_onyx | male | Pro |
| Puck | am_puck | male | Pro |
| Santa | am_santa | male | Pro |
| British English (8) | |||
| Alice | bf_alice | female | Pro |
| Emma | bf_emma | female | Free |
| Isabella | bf_isabella | female | Pro |
| Lily | bf_lily | female | Pro |
| Daniel | bm_daniel | male | Free |
| Fable | bm_fable | male | Pro |
| George | bm_george | male | Pro |
| Lewis | bm_lewis | male | Pro |
| Spanish (3) | |||
| Dora | ef_dora | female | Pro |
| Alex | em_alex | male | Pro |
| Santa | em_santa | male | Pro |
| French (1) | |||
| Siwis | ff_siwis | female | Pro |
| Hindi (4) | |||
| Alpha | hf_alpha | female | Pro |
| Beta | hf_beta | female | Pro |
| Omega | hm_omega | male | Pro |
| Psi | hm_psi | male | Pro |
| Italian (2) | |||
| Sara | if_sara | female | Pro |
| Nicola | im_nicola | male | Pro |
| Japanese (5) | |||
| Alpha | jf_alpha | female | Pro |
| Gongitsune | jf_gongitsune | female | Pro |
| Nezumi | jf_nezumi | female | Pro |
| Tebukuro | jf_tebukuro | female | Pro |
| Kumo | jm_kumo | male | Pro |
| Portuguese (3) | |||
| Dora | pf_dora | female | Pro |
| Alex | pm_alex | male | Pro |
| Santa | pm_santa | male | Pro |
46 Kokoro-82M voices. EasyVoice’s full catalog is 56 — these 46 plus 10 Modern Standard Arabic voices on a dedicated engine at /text-to-speech-arabic.
Kokoro TTS — frequently asked questions
What is Kokoro-82M?▾
Kokoro-82M is an open-weights neural text-to-speech model with 82 million parameters, released under the Apache 2.0 license. Despite being small enough to run on modest hardware, it produces natural, broadcast-intelligible speech that rivals commercial TTS engines. On EasyVoice it powers 46 voices across 8 languages: American and British English, Spanish, French, Hindi, Italian, Japanese, and Portuguese.
Can I use Kokoro TTS online without installing anything?▾
Yes. EasyVoice runs Kokoro-82M on its own GPUs, so you use it entirely in the browser — type text, pick a voice, download the MP3. There is no Docker container, no Python environment, no model download, and no GPU required on your side. You can generate 2,000 characters per day anonymously with no signup, and 5,000 per day with a free account.
Is Kokoro TTS free?▾
The model itself is free and open (Apache 2.0), and EasyVoice's free tier lets you use it at no cost: 2,000 characters/day with no signup, or 5,000/day with a free account — no credit card either way. Unlimited characters cost $9.99/month flat. Self-hosting is 'free' only in license terms; you still pay for the GPU and your own time.
Kokoro-82M vs a metered TTS API — which is cheaper?▾
It depends on what you need. If you only want a raw API endpoint and handle your own retries, storage, and tooling, DeepInfra serves Kokoro-82M at $0.62 per 1M characters — genuinely the cheapest metered option, and cheaper than EasyVoice below roughly 16M characters/month of pure API use. If you want the model plus a studio: long-form 50,000-character generations, MP3/WAV downloads, voice previews, projects, and an OpenAI-compatible API, EasyVoice is $9.99/month flat with no meter at all. Closed APIs like OpenAI tts-1 ($15/1M) cost 24x DeepInfra's Kokoro rate per character.
What does self-hosting Kokoro actually cost?▾
Kokoro-82M is small enough to run on a CPU for short clips, but long-form narration at usable speed wants a GPU. Rented GPUs on community GPU clouds (RTX 4090-class cards) run about $0.20–$0.60/hour in 2026 — an always-on server is $146–$438/month before you spend a single hour on CUDA drivers, model updates, audio storage, or uptime monitoring. Self-hosting makes sense for engineers who enjoy it or need on-premise audio; for everyone else, hosted Kokoro at $9.99/month flat is cheaper than the electricity conversation.
How fast is Kokoro TTS on EasyVoice?▾
About 1.1 milliseconds per character on EasyVoice's GPUs (own measured production data, July 2026). Short texts start returning audio in roughly one second warm. A 25,000-character script — a 20–30 minute YouTube narration — renders in about 28 seconds, and the Pro studio accepts up to 50,000 characters in a single generation (~55 minutes of audio).
Can I use Kokoro-generated audio commercially?▾
Yes. Kokoro-82M is Apache 2.0-licensed, and EasyVoice grants full commercial usage rights to generated audio on every plan, including the free tier — monetized YouTube videos, paid client work, courses, podcasts, and software products, with no attribution and no royalties.
Does Kokoro support Arabic?▾
No — Kokoro-82M does not include Arabic. EasyVoice covers Arabic with 10 dedicated Modern Standard Arabic voices on a separate purpose-built neural engine, which is why the full catalog is 56 voices across 9 languages: 46 Kokoro-82M voices plus 10 Arabic MSA voices, all under the same $9.99/month flat plan.
Is there a Kokoro API on EasyVoice?▾
Yes — an OpenAI-compatible one. POST to https://easyvoice.ae/api/v1/audio/speech with a Bearer key and an 'input' field; if your code already speaks the OpenAI TTS shape, switching is a one-line base-URL change. Free accounts get a free API key (5,000 chars/day, 12 voices); Pro unlocks all 56 voices.
No install, no signup, no card — 2,000 chars/day anonymous, 5,000/day with a free account