Zen3 Audio
Collection
Speech recognition + text-to-speech. • 7 items • Updated
How to use zenlm/zen-3-tts-custom-voice with Transformers:
# Use a pipeline as a high-level helper
from transformers import pipeline
pipe = pipeline("text-to-speech", model="zenlm/zen-3-tts-custom-voice") # Load model directly
from transformers import AutoModelForSeq2SeqLM
model = AutoModelForSeq2SeqLM.from_pretrained("zenlm/zen-3-tts-custom-voice", dtype="auto")Zen3 TTS variant for custom voice cloning from a short reference sample. Preserves speaker identity, prosody, and emotion across long-form synthesis in the Zen3 audio family.
Served at api.hanzo.ai as zen3-tts-custom-voice.
Mirror staging in progress. Use the canonical base via the base_model link in frontmatter until weights land here.
Built on Qwen/Qwen3-TTS-12Hz-1.7B-CustomVoice (Apache-2.0). Mirrored here for the Zen3 audio family.
Base model
Qwen/Qwen3-TTS-12Hz-1.7B-CustomVoice