DiffuPlex is a rolling masked diffusion framework designed to accelerate full‑duplex spoken dialog models by predicting multiple future user and assistant frames in a single backbone pass. It consumes only a confident prefix of each predicted future while maintaining the original frame rate, and revises unplayed future content when the interaction diverges. Two inference policies—DiffuPlex‑LISTEN and DiffuPlex‑SPEAK—offer speedups of 1.46×–1.59× in deployment‑path wall‑clock time and 1.61×–1.80× in Core LM speed, with all backbone invocations completing within the 80 ms interaction interval, while preserving conversational quality and, in the LISTEN policy, speech naturalness.
arXiv:2606. 11386v1 Announce Type: cross Abstract: Full-duplex spoken language models (FD-SLMs) enable seamless speech interaction by allowing models to listen and speak simultaneously, yet the internal mechanism by which they coordinate listening and speaking remains underexplored.
arXiv:2610.32349v1 Announce Type: new Abstract: Full-duplex speech models listen and speak at the same time, which makes them promising for natural, open-ended conversation. But can they sustain a co...
arXiv:2610.34991v1 Announce Type: new Abstract: End-to-end full-duplex spoken dialogue models enable natural, real-time spoken interactions with interruptions, backchannels, and overlapping speech. P...
The paper introduces llmovoice, a middleware that explicitly models voice context for large language model (LLM) serving in voice AI applications. By incorporating speaking rate, background noise, packet loss, and other paralinguistic factors into a bounded context, llmovoice guides the LLM to generate more aligned responses. Experiments show significant reductions in speaking‑rate errors, false interruptions, and model usage costs, especially in long voice sessions.
arXiv:2606. 07547v1 Announce Type: cross Abstract: Speech-based large language models are typically constrained to spoken replies, which limits their user-facing outputs to what can be verbalized and suppresses text-native capabilities such as code generation, structured analysis, and multi-step reasoning in realtime interaction, for tasks that require persistent, structured, and inspectable intermediate outputs.
arXiv:2608. 07631v1 Announce Type: cross Abstract: LLM-based full-duplex voice services allow users to speak while the assistant is responding.
Scalable Context Orchestration for Serving LLMs Over Voice presents llmovoice, a middleware that explicitly models voice context—including speaking rate, background noise, and packet loss—to guide large language model responses. By constructing a bounded voice context at each turn, llmovoice improves alignment with user preferences and reduces errors, achieving a 52.4% drop in speaking‑rate alignment error and a 0.9% false‑interruption rate under packet loss. In addition, it cuts model usage costs dramatically, lowering per‑turn cost by up to 24.9× while maintaining 98.7% of baseline answer quality in long sessions.
arXiv:2610.17124v1 Announce Type: cross Abstract: Full-duplex spoken language models (FD-SLMs) support simultaneous listening and speaking, but often lag behind half-duplex spoken language models (HD...
The study evaluates conversational speech models by placing them at 427 pauses and 482 turn ends from 176 unscripted conversations, scoring their responses on timing, latency, length, wording, and recall of recent dialogue. Hosted APIs (GPT Realtime, Gemini 3.1 Flash Live, Grok Voice) respond at almost every pause, with latency ranging from 1–4 s for silence-based endpointers and about 8 s for semantic turn detection, while full‑duplex models (PersonaPlex, Moshi) remain silent more often and can interrupt within human reaction time. Recall rates vary: hosted systems remember 41–60 % of facts under a Claude Opus 5.5 judge and 19–26 % with a strict word matcher; full‑duplex models recall 17–22 % and 12–15 % respectively.
Full-duplex speech models require training data that preserves turn-taking, overlap, interruption, and backchannel behavior, yet these signals are entangled across speakers in noisy real-world recordi...
The paper introduces a frontend‑backend architecture for full‑duplex speech‑to‑speech models that enables tool calls while preserving natural conversational flow. The frontend emits a delegation token and streams ASR transcripts to a text‑based backend LLM, which performs tool calls and returns results that are re‑injected into the frontend via a lightweight prefill‑and‑repeat mechanism before streaming TTS synthesis. In single‑turn evaluations the system achieves 92‑97% tool‑call recall, 81.2% accuracy in rejecting irrelevant calls, and competitive performance on Full‑Duplex‑Bench‑V3 and EVA‑Bench when paired with a large backend model.
DuplexLedger is a bilingual benchmark comprising 1,094 persistent sessions that align four full‑duplex scenarios with event‑level Write/Update/Ignore/Retain supervision, explicit pre/post memory states, and 6,564 same‑session questions and checks. Evaluation of five full‑duplex models shows an interaction‑memory gap, with no system achieving strong boundary control and coherent final‑state memory; the best final‑state fact coverage is only 2.01%. The authors propose ACT‑Mem, an 8.92M‑parameter sidecar that improves latest‑value accuracy from 4.11% to 8.68% on held‑out sessions while reducing stale assertions and distractor adoption, though it slows interruption stopping.