Skip to content

Benchmark

OmniGAIA

Scores are percentages. Higher is better.

Models and systems measured
6
Scores recorded
9
Sources
2
IQ weight
0.12

Score distribution

59.9%
Mean
57.2%
Median
13.6%
Std dev
49.8%
10th percentile
74%
90th percentile
33.9%
Min
33.9%78.6%
The band covers the middle 80% of scores. The line marks the median.

This benchmark has a weight of 0.12 for chat models, measured over 5 models. The weight is how much it counts toward llminfo IQ. We measure it from the data, not from the raw scores.

Leaderboard

#Model or systemScore
1Gemini 3.8 FlashWeb Search · 无 Harness· Reported by Qwen· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.78.6%
2Qwen3.8 Omni Flash· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim74%
3Qwen3.8 Omni FlashWeb Search · 无 Harness· Reported by Qwen· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.74%
4Qwen3.5 Omni PlusQwen3.5-Omni release · no thinking prompt · no <answer> formatting · DeepSeek-V3.2-Thinking judge· Reported by Qwen, Qwen Team· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.57.2%
5Qwen3.5 Omni PlusWeb Search · 无 Harness· Reported by Qwen· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.57.2%
6MiMo V2 Omni· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim49.8%
7MiMo V2 OmniXiaomi MiMo-V2-Omni release table · multimodal perception and reasoning · harness unspecified· Reported by Xiaomi MiMo· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.49.8%
8Qwen3.5 Omni FlashQwen3.5-Omni release · no thinking prompt · no <answer> formatting · DeepSeek-V3.2-Thinking judge· Reported by Qwen, Qwen Team· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.33.9%

Not ranked

The source marks these results as descriptive, not capability scores, so we list them by name instead of ranking them.

Model or systemScore
Seed-2.0-LiteWeb Search · 无 Harness· Reported by Qwen· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.64.4%

Data credits: BenchBoard by SYuan03 (github.com/SYuan03/benchboard). MIT covers its code and data organization; no results licence is inferred. Numeric relays retain the original stating-source links. · Data by LLM Stats