VoiceGiraffe (Cascaded Captions): leaderboard

VoiceGiraffe (2026) tests hour-scale audio understanding with 1,500 four-option questions on long English and Chinese recordings. Under the cascaded-caption protocol the model captions the audio in 30-second windows and then answers from its own concatenated captions; this board is accuracy in percent over all questions, higher is better.

Metric: Overall accuracy (%) over the 1,500 four-option VoiceGiraffe questions on 123 English and Chinese recordings of 30 minutes to 2 hours; cascaded caption aggregation: the model captions the recording in 30-second sliding windows (speech, speaker traits, sound events, music) and then answers from the concatenated captions and the question itself, without an external reasoning model; answers are matched by regular expression; higher is better. Source: arxiv.org. Saturation forecast: Around February 2027. 12 models tracked. Updated 7 Oct 2026.

Top models

#ModelSettingScoreOverall rank
1Gemini 3.1 Pro (Preview)75.1#102
2Gemini 2.5 Pro72.6#280
3Qwen 3.5 Omni Plus66.2#261
4Qwen3.5 Omni Flash57.9#610
5Qwen3 Omni 30B A3B Instruct50.6#928
6MOSS-Audio-8BThinking45.3#2138
7Audio Flamingo 344.5
8Qwen 3 Omni 30B A3B (Thinking)44.2
9Qwen2.5-Omni-7B41.1#1888
10MiMo-Audio-7B38.2
11MOSS-Audio-8B-Instruct36.7#2018
12Phi-4 Multimodal Instruct30.1#4155

No result here: #2 Claude Fable 5.2, #3 Gemini 4 Neon, #4 GPT-6.2 Astra.

Interactive version: theaggregate.ai/benchmark?slug=voicegiraffe-cascaded-captions · How It Works · Data refreshed daily, snapshot 2026-11-22.