Audio Flamingo 3: Advancing Audio Intelligence with Fully Open Large Audio Language Models

Sreyan Ghosh, Arushi Goel, Jaehyeon Kim, Sonal Kumar, Zhifeng Kong, Sang-gil Lee, Chao-Han Huck Yang, Ramani Duraiswami, Dinesh Manocha, Rafael Valle, Bryan Catanzaro

NVIDIA· University of Maryland

Large Audio Language Models; Audio and Music Understanding; Multi-Dialogue Chat; Reasoning

⋅ NeurIPS ⋅ Project Page ⋅OpenReview

Abstract

We present Audio Flamingo 3 (AF3), a fully open state-of-the-art (SOTA) large audio-language model that advances reasoning and understanding across speech, sound, and music.

AF3 introduces:

AF-Whisper, a unified audio encoder trained using a novel strategy for joint representation learning across all 3 modalities of speech, sound, and music;
flexible, on-demand thinking, allowing the model to do chain-of-thought-type reasoning before answering;
multi-turn, multi-audio chat;
long audio understanding and reasoning (including speech) up to 10 minutes; and
voice-to-voice interaction.

To enable these capabilities, we propose several large-scale training datasets curated using novel strategies, including AudioSkills-XL, LongAudio-XL, AF-Think, and AF-Chat, and train AF3 with a novel five-stage curriculum-based training strategy. Trained on only open-source audio data, AF3 achieves new SOTA results on over 20+ (long) audio understanding and reasoning benchmarks, surpassing both open-weight and closed-source models trained on much larger datasets.