Paper 2604.15804
Qwen3.5-Omni Technical Report
- Published
- Apr 2026
- Research lab
- Independent
- Citations
- 87
- GitHub
- Not linked
01 In brief
Summary
Qwen3.5-Omni is a fully omnimodal large language model that scales to hundreds of billions of parameters and supports a 256k context length.
It is pretrained on a massive dataset including over 100 million hours of audio-visual content.
The model uses a Thinker-Talker architecture with Hybrid-Attention Mixture-of-Experts (MoE) for both components, enabling efficient long-sequence inference.
Key innovations include ARIA (Adaptive Rate Interleave Alignment) for stable streaming speech synthesis, a multi-codebook codec for single-frame synthesis, and explicit timestamps for improved temporal perception.
Qwen3.5-Omni supports 113 languages/dialects for speech recognition and 36 for speech synthesis, with zero-shot voice cloning.
It achieves state-of-the-art results across 215 audio and audio-visual benchmarks, surpassing Gemini-3.1 Pro in key audio tasks and matching it in comprehensive audio-visual understanding.
The model also demonstrates new capabilities such as controllable audio-visual captioning, real-time interaction with semantic interruption, and Audio-Visual Vibe Coding, where it generates executable code from audio-visual instructions.
The model is available in Plus and Flash variants, with first-packet latencies of 435ms and 235ms for audio input, respectively.
02 From the paper
Abstract
In this work, we present Qwen3.5-Omni, the latest advancement in the Qwen-Omni model family. Representing a significant evolution over its predecessor, Qwen3.5-Omni scales to hundreds of billions of parameters and supports a 256k context length. By leveraging a massive dataset comprising heterogeneous text-vision pairs and over 100 million hours of audio-visual content, the model demonstrates robust omni-modality capabilities. Qwen3.5-Omni-plus achieves SOTA results across 215 audio and audio-visual understanding, reasoning, and interaction subtasks and benchmarks, surpassing Gemini-3.1 Pro in key audio tasks and matching it in comprehensive audio-visual understanding. Architecturally, Qwen3.5-Omni employs a Hybrid Attention Mixture-of-Experts (MoE) framework for both Thinker and Talker, enabling efficient long-sequence inference. The model facilitates sophisticated interaction, supporting over 10 hours of audio understanding and 400 seconds of 720P video (at 1 FPS). To address the inherent instability and unnaturalness in streaming speech synthesis, often caused by encoding efficiency discrepancies between text and speech tokenizers, we introduce ARIA. ARIA dynamically aligns text and speech units, significantly enhancing the stability and prosody of conversational speech with minimal latency impact. Furthermore, Qwen3.5-Omni expands linguistic boundaries, supporting multilingual understanding and speech generation across 10 languages with human-like emotional nuance. Finally, Qwen3.5-Omni exhibits superior audio-visual grounding capabilities, generating script-level structured captions with precise temporal synchronization and automated scene segmentation. Remarkably, we observed the emergence of a new capability in omnimodal models: directly performing coding based on audio-visual instructions, which we call Audio-Visual Vibe Coding.