Audio Flamingo 3: Advancing Audio Intelligence with Fully Open Large Audio Language Models
About
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: (i) AF-Whisper, a unified audio encoder trained using a novel strategy for joint representation learning across all 3 modalities of speech, sound, and music; (ii) flexible, on-demand thinking, allowing the model to do chain-of-thought-type reasoning before answering; (iii) multi-turn, multi-audio chat; (iv) long audio understanding and reasoning (including speech) up to 10 minutes; and (v) 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.
Related benchmarks
| Task | Dataset | Result | Rank | |
|---|---|---|---|---|
| Automatic Speech Recognition | LibriSpeech clean (test) | WER1.57 | 1207 | |
| Automatic Speech Recognition | LibriSpeech (test-other) | WER3.13 | 1206 | |
| Audio Captioning | AudioCaps (test) | CIDEr0.79 | 157 | |
| Automatic Speech Recognition | LibriSpeech Other | WER3.13 | 123 | |
| Speaker Verification | VoxCeleb1 (Vox1-O) | -- | 105 | |
| Multimodal Understanding | MMMU | MMMU Score72.42 | 102 | |
| Audio-Visual Question Answering | AVQA | Accuracy64.3 | 85 | |
| Audio Captioning | AudioCaps | CIDEr70 | 66 | |
| Music Genre Classification | GTZAN | Accuracy83.2 | 62 | |
| Audio-visual understanding | Daily-Omni | Accuracy52.5 | 58 |