Apriel Voice 27B unifies speech reasoning and tool use
Original title: Apriel Voice 27B Preview: An Open Model for Speech, Reasoning, and Tool Use
Apriel Voice 27B Preview processes speech inputs and generates responses with tool calls on a shared backbone built from Gemma 4 26B-A4B.
Apriel Voice is an open speech-language model that combines speech understanding with reasoning and tool calling. It uses two-stage training to adapt the base model for speech interface and joint optimization. The model achieves strong performance on audio benchmarks including 98.90 on Big Bench Audio and supports interactive voice interactions on a custom platform.
- Scores 98.90 on Big Bench Audio and 81.13 on Speech-BFCL
- 65.51% task success rate on τ-Voice with 1.91s mean latency
- 51.27% English EVA accuracy and 74.13% experience pass@1
From the original article
1 min read at huggingface.co
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.
By Aman Tiwari, Akshay Kalkunte, Akintunde Oladipo · huggingface.co
A short excerpt; the article and its images belong to the publisher. Summary, key facts and assessment above are AI-written from the article and can be wrong.