# Examples & Models

Kokoro, Qwen, F5, Higgs, Chatterbox, and Fish Speech are the supported speech backends. Each model narrates the identical excerpt from Steve Jobs's [Stanford Commencement Address](https://news.stanford.edu/2005/06/14/jobs-061505/) so you can directly compare audio timbre, pacing, pronunciation, and latency.

## Speech Model Comparison

Compare five open model architectures running on Apple Silicon (MLX) and AWS Batch:

### Kokoro-82M
by Hexgrad
82M params

The Steve Jobs commencement speech excerpt, spoken with Kokoro-82M on Apple Silicon (MLX) or PyTorch.

StyleTTS 2 + ISTFTNet • 24 kHz • Apache-2.0

[Listen & Compare →](/examples/basic)

### Qwen3-TTS
by Alibaba Qwen
0.6B params

The same commencement speech excerpt, spoken with Qwen3-TTS 0.6B CustomVoice with the Ryan preset.

CustomVoice 12Hz Transformer • 24 kHz • Apache-2.0

[Listen & Compare →](/examples/qwen)

### F5-TTS
by SWivid
Flow matching

The same commencement speech excerpt, spoken with non-autoregressive F5-TTS via MLX on Apple Silicon.

Non-autoregressive Flow Matching • 24 kHz • MIT

[Listen & Compare →](/examples/f5)

### Higgs Audio v3
by Boson AI
Boson AI 4B

The same commencement speech excerpt, spoken with Boson AI's expressive 4B Higgs Audio v3 via MLX on Apple Silicon.

Higgs Audio 4B Transformer • 24 kHz • Research / Non-Commercial

[Listen & Compare →](/examples/higgs)

### Chatterbox-TTS
by Resemble AI
520M Hybrid

The same commencement speech excerpt, spoken with Resemble AI's hybrid LLaMA-520M and Matcha-TTS flow matching.

LLaMA-520M + Matcha-TTS Flow • 24 kHz • Apache-2.0

[Listen & Compare →](/examples/chatterbox)

### Fish-Speech
by Fish Audio
Dual-AR

The same commencement speech excerpt, spoken with Fish Audio's dual-autoregressive model with natural prosody and breathing.

Dual-AR Transformer + VQ-GAN • 44.1 kHz • CC-BY-NC-SA-4.0

[Listen & Compare →](/examples/fish)

## Embed Features & Theming

Customizing player appearance and article content extraction:

- [**Themed Player** CSS variables ](/examples/themed) The same commencement speech excerpt with Kokoro, styled via CSS custom properties (--auritus-*).

- [**Ignore & Pronounce Markup** Article parsing ](/examples/ignore) John Stuart Mill's On Liberty demonstrating data-auritus-ignore to skip ads and data-auritus-pronounce.