Get started with Hutash
Three steps from download to your first generation. Everything runs on your own machine.
Get started with Hutash
Three steps from download to your first generation. Everything runs on your own machine.
Install: Three steps to your first generation
Step 1 — Download and install
Get the Windows installer from hutash.com/download and run it. No account, no sign-in, no licence key.
Step 2 — Install a model
Open the marketplace and click Install on any model. The marketplace checks your hardware first, so you know what will run before you download. From a terminal:
hutash install kokoro
Step 3 — Generate
Open Studio, type your text, and click Generate. Output is written straight to your disk. From a terminal:
hutash run kokoro tts --text "Hello world"
Requirements: System requirements
| Component | Minimum | Recommended |
|---|---|---|
| Operating system | Windows 10 (64-bit) | Windows 11 |
| RAM | 8 GB | 16 GB or more |
| GPU | None — CPU inference works | NVIDIA GPU with 8 GB VRAM or more |
| Disk | 2 GB for the app and one small model | 30 GB to run the full v1.0 catalogue |
| Network | Needed to download models | Not needed for generation |
Linux and macOS support are coming soon.
Models: What you can install
Grouped by what they do. The models page lists sizes, hardware and licences in full.
Building your own
Packaging a model or an app of your own? Developer Mode lets you load a local package and test it before you publish anything.
| What it does | Models |
|---|---|
| Voiceover | Indic Parler TTS, Kokoro, Parler TTS Mini, Svara TTS |
| Music | ACE-Step 1.5 |
| Clone Voice | Chatterbox, Dia 1.6B |
| Transcribe | Distil-Whisper, Whisper Large V3 |
| Translate | MADLAD-400 3B, Opus-MT English-Hindi |
| Chat | Gemma 4 E4B, Qwen3 1.7B, Qwen3.5 4B |