Metraipolis › archive › dc_e71810acc3ee

Apriel Voice 27B Preview: An Open Model for Speech, Reasoning, and Tool Use — arXiv:2610.40738

Source
arXiv (en)
Original
https://arxiv.org/abs/2610.40738
License (as archived)
CC0-1.0
Archived
2026-11-03T08:05:39.108Z
SHA-256 of the archived text
edde095780a02303e66ed0a1030086ac4deffd4ed4005115f40f4f8345286773
Stable URL of this page
https://metraipolis.com/documento/dc_e71810acc3ee

How to cite

Apriel Voice 27B Preview: An Open Model for Speech, Reasoning, and Tool Use — arXiv:2610.40738 — arXiv (en), https://arxiv.org/abs/2610.40738 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_e71810acc3ee (Metraipolis archive dc_e71810acc3ee, sha256 edde095780a02303e66ed0a1030086ac4deffd4ed4005115f40f4f8345286773)
BibTeX
@misc{dce71810acc3ee,
  title = {Apriel Voice 27B Preview: An Open Model for Speech, Reasoning, and Tool Use — arXiv:2610.40738},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2610.40738},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_e71810acc3ee; sha256 edde095780a02303e66ed0a1030086ac4deffd4ed4005115f40f4f8345286773},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2610.40738. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

A capable voice assistant must do more than recognize words; it must understand requests, reason over context, sustain conversations, and act through tools. We present Apriel Voice 27B Preview, an open speech-language model that unifies these capabilities within a shared reasoning backbone while retaining text and image input modalities. Built on the post-trained Gemma 4 26B-A4B checkpoint, Apriel Voice directly processes speech and generates both responses and tool calls without requiring a separate dialogue model. We introduce a two-stage training procedure to adapt the model to speech. The first stage establishes the speech interface, while the second jointly adapts the projector and backbone through supervised fine-tuning, text replay, and source-dependent cross-entropy and forward Kullback--Leibler divergence supervision. Automatically constructed speech-conditioned dialogues provide conversational training data, while local teacher serving, overlapping teacher and student computation, and fused full-vocabulary losses enable efficient online distillation. In thinking mode, Apriel Voice achieves a score of 98.90 on Big Bench Audio, 81.13 on Speech-BFCL, and a task success rate of 65.51% on $\tau$-Voice, with a mean latency of 1.91 seconds. It also achieves 51.27% English EVA accuracy and 74.13% experience pass@1. Finally, we deploy Apriel Voice 27B Preview on a custom serving platform to support interactive voice interactions.

Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.CL, nuovi lavori del giorno).
Categorie arXiv: Sound (cs.SD) ; Computation and Language (cs.CL); Machine Learning (cs.LG); Audio and Speech Processing (eess.AS).
Pagina del lavoro: https://arxiv.org/abs/2610.40738 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito).
Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina.
Letto il 2026-11-03T07:59:53.609Z dalla pagina https://arxiv.org/list/cs.CL/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).