Higgs Audio v3 TTS (4B) — transformers trust remote code port A trust remote code packaging of bosonai/higgs audio v3 tts 4b that loads with plain 🤗 transformers (no SGLang). The weights are the original checkpoint, copied unchanged; only a small modeling .py / configuration .py pair and an auto map were added. The model is a standard Qwen3 4B backbone plus a fused multi codebook audio embedding/head. Reference audio encoding and waveform decoding use the transformers native bosonai/higgs audio v2 tokenizer ( higgs audio v2 tokenizer ), loaded automatically on first use. Requires transformers = 5.5 . Usage generate speech returns a mono 24 kHz waveform as a CPU float32 tensor [L] . Notes Generation uses Higgs' delay pattern across 8 codebooks (vocab 1026, incl. BOC/EOC specials); de delay + decode are handled internally. The codec runs in fp32 (decode is unstable in bf16); the LM backbone runs in the dtype you load it in (bf16 recommended). License: research/non commercial, inherited from the upstream checkpoint — see LICENSE .
We use cookies for essential functionality and analytics. You can accept or reject analytics cookies.Cookie policy