Qwen3-TTS-12Hz-1.7B-CustomVoice vs higgs-tts-2-3b-base

Specs, VRAM requirements and download trends — updated 21 October 2026.

Qwen · A · Open weights
Params
1.9 B
Context
—
30d
2.0 M
apache-2.0 · commercial OK
bosonai · B · Open weights
Params
5.8 B
Context
2,048
30d
415 K

Specification comparison

Differences are highlighted; identical values are muted.

Specification Qwen3-TTS-12Hz-1.7B-CustomVoice higgs-tts-2-3b-base
Parameters 1.9 B 5.8 B
Architecture Qwen3TTSForConditionalGeneration HiggsAudioV2ForConditionalGeneration
Context length — 2,048
Licence apache-2.0 other
Commercial use Allowed Unknown
Languages — 4
Downloads 30d 2,038,337 414,882
Downloads all time 17.3 M 4.4 M
Quantizations on the Hub SAFETENSORS SAFETENSORS
Gated No No
First seen on the Hub 2026-01-21 2025-07-01

Download trend

Daily snapshots, last 87 days (28 Jul – 22 Oct)

2.6 M368 K
Qwen3-TTS-12Hz-1.7B-CustomVoice higgs-tts-2-3b-base

VRAM side by side

On RTX 4090 · 24 GB · 8K context unless noted

Qwen3-TTS-12Hz-1.7B-CustomVoice · fp16 5.8 GB / 24 GB
✅ Runs comfortably
higgs-tts-2-3b-base · fp16 14.1 GB / 24 GB
✅ Runs comfortably

Adjacent comparisons