English static mirror for SEO/GEO · AI-assisted translation · Read Chinese original

NVIDIA NemotronLabs VoiceChat 11B: First Open-Source Full-Duplex Speech Agent Base Model with Native Tool Calling

Forum topic · 小凯 · 2026-08-10

Summary

NVIDIA released NemotronLabs VoiceChat 11B on August 9 via Hugging Face as a research-grade foundation for full-duplex voice agents. It is the first open-source speech-to-speech model that can invoke external tools during conversation while remaining fully duplex. The architecture replaces the traditional ASR-LLM-TTS cascade with a single unified network: a Fast Conformer encoder (16 kHz streaming), a Nemotron Nano v2 9B LLM backbone, and an NVIDIA TTS decoder/codec (22.05 kHz output). Reported smooth turn-taking latency is 448 ms. Tool calls are emitted on a side channel in a JSON block, and a preset 'on-hold' utterance fills silence while APIs run. Benchmarks include Full-Duplex-Bench 1.0 and BFCL-v3 speech. Requirements: 80 GB single GPU, OpenMDW-1.1 research-only license, vLLM runtime, English-focused training data, ASCII-only system prompts, 2-minute audio context, and a 5-tool cap per session.

Overview

NVIDIA released NemotronLabs VoiceChat 11B on August 9 via Hugging Face as a foundation model for builders of full-duplex voice agents, not for end consumers. It is the first open-source speech-to-speech model that can invoke external tools during conversation without giving up duplex behavior.

Why It Matters

Previous open-source full-duplex stacks fell into one of three buckets: listen-only, speak-only, or speak-and-listen without tool calling. VoiceChat 11B closes the third gap by emitting <TOOLCALL> blocks on a side channel while the agent delivers a preset "on-hold" utterance during the 2-second wait for an API response.

Architecture

A single unified network replaces the ASR + LLM + TTS cascade:

  • Fast Conformer encoder from Nemotron-Speech-Streaming-En-0.6b, continuously encoding 16 kHz audio.
  • Nemotron Nano v2 9B as the LLM backbone, consuming audio tokens and predicting text tokens.
  • NVIDIA TTS decoder + codec producing 22.05 kHz audio.
  • Independent tool-calling output channel.
  • Training data totals ~550,000 hours of audio (real + synthetic), building on SALM-Duplex and Audio Flamingo 3 research.

    Tool Calling Design

    Tool calls use the format <TOOLCALL>[{"name":..., "arguments":...}]</TOOLCALL>. Business code returns results via <TOOL_RESPONSE>[...]</TOOL_RESPONSE>. Each tool can declare an on-hold utterance spoken at invocation, eliminating awkward silence while async calls resolve.

    Benchmarks

    Full-Duplex-Bench 1.0

  • Smooth turn-taking: TOR 0.82 / 448 ms
  • User-interruption: TOR 1.00 / 480 ms
  • Pause-handling (lower is better): synthetic 0.153, Candor 0.255
  • AU Harness BFCL-v3 (speech)

  • Simple: 58.5%
  • Multiple: 62.5%
  • Parallel: 42.5%
  • Parallel-multiple: 27.5%
  • Irrelevance: 89.6%
  • Average: 56.1%
  • Full-Duplex-Bench v3 tool calling

  • Tool selection: 82.5%
  • Argument accuracy: 44.2%
  • Pass@1: 33%
  • NVIDIA self-reports 2nd place among open-source full-duplex models on VoiceBench and Full-Duplex-Bench 1.0.

    Deployment Requirements & Limits

  • License: OpenMDW-1.1, research use only; commercial licensing is separate.
  • No hosted API. NGC container and Hugging Face weights only.
  • Hardware: single 80 GB GPU (A100 / H100 / H200 / B100 / B200 / RTX-6000), x86_64 Linux, vLLM runtime.
  • Context: 2 minutes of audio max; quality can degrade into unrecoverable garble after several turns.
  • Artifacts: occasional runaway self-talk after turn end; occasional dropped words in user STT.
  • Tool limits: 5 tools per session max; no parallel tool calls; user cannot interrupt during tool execution (to preserve <TOOLCALL>/<TOOL_RESPONSE> causality).
  • Prompts: system prompts and tool responses must be ASCII-only and TTS-friendly (no emoji, no full-width symbols, no Unicode quotes). The Fast Conformer is English-stream and training data is predominantly English, which limits Chinese-language support.
  • Key Takeaways

    1. It is a base model, not a product. 448 ms latency, 80 GB VRAM, and a research-only license position it as a baseline for voice-agent engineering teams and a research stimulus for competitors such as Moshi, Audio Flamingo 3, and SALM-Duplex. 2. On-hold utterances are an underrated design pattern. Filling the silence during async tool calls addresses a usability bottleneck distinct from raw latency. The same Agent-era paradigm shift appears in Cosmos 3's first-class action tokens and SeedRealtime's internalized "when to speak" modeling. 3. The H2 2026 full-duplex landscape is taking shape. Open-source bases provide tool calling and Chinese teams must navigate two portability concerns: OpenMDW's commercial boundary and the ASCII-only prompt constraint for Chinese interaction.

    Best Fit / Poor Fit

  • Best fit: Voice AI R&D teams with 80 GB GPU capacity (in-car voice, contact-center CX, game NPCs, retail kiosks, accessibility).
  • Poor fit: Consumer products needing immediate production deployment, businesses requiring deep Chinese support, and teams needing a hosted API for fast integration.
  • Sources

  • Official model card: https://huggingface.co/nvidia/NVIDIA-NemotronLabs-VoiceChat-11B
  • Official code (NeMo Speech branch): https://github.com/NVIDIA-NeMo/Speech/tree/nemotron-labs-voicechat
  • Official NGC container: https://catalog.ngc.nvidia.com/orgs/nim/nvidia/containers/nemotron-labs-voicechat
  • OpenMDW 1.1 license: https://github.com/OpenMDW/OpenMDW/blob/main/1.1/LICENSE.OpenMDW-1.1
  • MarkTechPost technical write-up: https://www.marktechpost.com/2026/08/09/nvidia-releases-nemotronlabs-voicechat-11b-an-open-full-duplex-speech-to-speech-model-with-450-ms-turn-taking-and-live-tool-calling
  • Full-Duplex-Bench 1.0 paper: https://arxiv.org/abs/2503.04721
  • AU Harness BFCL-v3: https://github.com/ServiceNow/AU-Harness
  • VoiceBench paper: https://arxiv.org/abs/2410.17196

Tags

#nvidia#voicechat-11b#full-duplex#speech-agent#tool-calling#open-source#nemotronlabs#voice-ai

This page is an English static mirror generated for search and AI citation. It may be a full translation or structured summary of the Chinese original. Canonical interactive discussion lives on the Chinese page: https://zhichai.net/topic/178630988