Omni-DeepSearch - Image-Text Search: leaderboard

Omni-DeepSearch (2026) tests audio-driven deep search: from one or more audio clips alone, a model must infer the audio entity, call text, image and video search tools and give a short verifiable answer, judged by a majority of three LLM judges. This board is accuracy on the 160 questions whose answer must be verified from searched images, in percent; higher is better.

Metric: Accuracy (%) on the 160 single-audio image-text search questions (the answer must be verified from image evidence), audio-driven deep search: starting from audio clips alone, the model calls text, image and video search tools (at most 10 text or image searches and 3 video searches per question) and gives a short verifiable answer, judged by majority vote of GPT-5.4, Gemini-3-pro and Claude-Sonnet-4.6; higher is better. Source: arxiv.org. Saturation forecast: Around February 2027. 12 models tracked. Updated 7 Oct 2026.

Top models

#ModelScoreOverall rank
1Gemini 3 Pro38.75#145
2Gemini 3 Flash21.25#175
3Qwen 3.5 Omni Plus15.62#254
4Gemini 2.5 Pro15#276
5MiMo-V2.514.38#383
6MiMo-V2-Omni11.88#465
7Qwen 3 Omni 30B A3B (Thinking)10.62
8Qwen3.5 Omni Flash6.88#602
9Qwen3 Omni 30B A3B Instruct1.88#930
10Qwen2.5-Omni-7B0.62#1884
11Qwen2.5-Omni-3B0.62#3519
12Gemini 2.5 Flash Lite0#964

No result here: #1 Claude Fable 5.2, #3 Gemini 4 Neon, #4 GPT-6.2 Astra.

Interactive version: theaggregate.ai/benchmark?slug=omni-deepsearch-image-text-search · How It Works · Data refreshed daily, snapshot 2026-11-21.