ChatTTS 2026: 39.3k-Star Open-Source Dialogue TTS with Laughter, Pauses, and Token-Level Prosody Control

Most open-source TTS in 2026 still sounds like “1990s GPS narrator with extra reverb.” ChatTTS is the first widely-adopted exception — a 39.3k-star generative speech model specifically trained for dialogue, not narration, with token-level control over laughter, pauses, interjections, and prosody that finally crosses the “doesn’t make you wince” threshold.

If you’re building voice agents, AI podcasts, multi-character TTS for games, or any voice product where flat narration kills the experience — ChatTTS is the default open-source pick in 2026.

TL;DR #

• What: Open-source dialogue-optimized TTS by 2noise (Chinese + English)
• GitHub: 39.3k stars
• License: AGPL-3.0 (code) + CC BY-NC 4.0 (model) — non-commercial use only on the open weights
• VRAM: 4 GB minimum for 30-second clips; comfortable at 8 GB
• Speed: RTF (Real-Time Factor) ~0.3 on RTX 4090 (faster than real-time)
• Training data: 40,000+ hours open-source / 100,000+ hours full model

1. Why ChatTTS Beats Traditional TTS for Dialogue #

The legacy split:

• Concatenative TTS (festival, etc.) — mechanical, no prosody, dying
• Neural TTS (Tacotron / FastSpeech / VITS) — fluid but monotone, optimized for narration
• Commercial APIs (ElevenLabs / OpenAI TTS) — natural but $0.18-0.50/1000 chars and closed

ChatTTS sits in a new fourth category: autoregressive generative TTS with explicit prosodic control tokens. You don’t just type text — you can mark up [laugh], [uv_break] (umm), [lbreak] (long pause), and the model produces a vocal performance, not just speech.

For dialogue use cases (voice agents, AI podcasts, NPC dialogue in games), this is the difference between “obvious robot” and “could be a real person on a bad phone line.” For narration, classical neural TTS is still often better.

2. Hardware Requirements (Realistic Numbers) #

For self-hosted production, the entry point is a $0.30-0.50/hr GPU cloud (Vast.ai, RunPod, or a DigitalOcean GPU droplet ) — much cheaper than ElevenLabs at any meaningful volume.

3. Quick Install (10 minutes on a GPU machine) #

git clone https://github.com/2noise/ChatTTS
cd ChatTTS
pip install -r requirements.txt
# Or via pip:
pip install ChatTTS

Hello world:

import ChatTTS
import torchaudio
import torch

chat = ChatTTS.Chat()
chat.load(compile=False)  # Set True for ~30% speedup after first run

texts = ["Hello, this is a dialogue TTS test [uv_break] does it sound natural?"]
wavs = chat.infer(texts)

torchaudio.save("out.wav", torch.from_numpy(wavs[0]), 24000)

First run downloads ~2 GB of model weights. Subsequent runs are instant.

4. The Prosody Control Tokens (the killer feature) #

The reason ChatTTS feels alive — these tags work mid-text:

Example:

text = "So I told him [uv_break] there's no way that's true [laugh] [lbreak] but he kept insisting."
wavs = chat.infer([text])

This is what closes the “robot vs human” gap. Use sparingly; over-tagging sounds rehearsed.

5. Multi-Speaker — Stable Voices Across Sessions #

ChatTTS generates a different “speaker” each invocation by default. For consistent characters (NPC voice, persistent agent personality), seed a speaker once and reuse:

# Generate and save a stable speaker
rand_spk = chat.sample_random_speaker()
torch.save(rand_spk, "speaker_alice.pt")

# Use in subsequent runs
spk = torch.load("speaker_alice.pt")
params_infer_code = ChatTTS.Chat.InferCodeParams(spk_emb=spk)
wavs = chat.infer(texts, params_infer_code=params_infer_code)

Pattern: pre-generate 5-10 distinct speaker embeddings during setup. Pick the one matching each character. Voice stays stable across the entire production.

6. Licensing Caveat (Read This Before Production) #

Two licenses, two different obligations:

• Code: AGPL-3.0 — copyleft, derivative works must open-source under AGPL
• Model weights: CC BY-NC 4.0 — non-commercial use only

For commercial production: contact 2noise for commercial model licensing, OR fine-tune your own model from scratch on the open ChatTTS architecture (significant effort but legally clean).

For hobby, research, internal tooling, and most agent prototyping: the NC license is fine.

This is the only friction point on adoption. If your product directly monetizes generated voices, factor in licensing cost (or pick commercial-friendly alternatives like Coqui XTTS-v2).

7. Production Pattern #

For agent voice / podcast pipeline:

   Text input (from LLM agent / script generator)
            │
            ▼
   Light preprocessing (add prosody tags via rules)
            │
            ▼
   ChatTTS inference (GPU-backed FastAPI service)
            │
            ▼
   Output audio (WAV / Opus / streaming)
            │
            ▼
   Optional post-processing (loudness normalize, denoise)

Stand it up on a GPU-equipped HTStack Hong Kong VPS with GPU or a Vast.ai instance, expose via FastAPI, and your stack has voice for ~$0.001 per minute generated (vs ElevenLabs at ~$0.30/min).

8. When to Use ChatTTS vs Alternatives #

9. Pitfalls #

1. Over-tagging prosody — sprinkling [laugh] and [uv_break] everywhere sounds rehearsed. Less is more
2. Forgetting to seed speakers — every call without spk_emb is a different voice. Always pre-generate
3. Running on CPU and complaining about speed — RTF is 30-100× worse without GPU. Just rent a GPU
4. Ignoring the NC license — using ChatTTS for paid voice products is a legal risk

TL;DR #

ChatTTS = first open-source TTS that handles dialogue convincingly. 39.3k stars, 4 GB VRAM minimum, prosody-aware via control tokens, stable speakers via embedding seeds. Use for voice agents, AI podcasts, NPCs. Watch the CC BY-NC license for commercial use.

Spin up a GPU instance, run the 10-line install in section 3, and you’ll hear in 5 minutes why this displaced every other open-source TTS in the discussion.

Part of dibi8’s multi-modal content stack — see the upcoming Multi-Modal Content Pipeline collection for ChatTTS + Whisper + Stable Diffusion + ComfyUI as a full audio/visual creator pipeline.