What is the RefCOCO-avg leaderboard?

The RefCOCO-avg leaderboard ranks 7 AI models based on their performance on this benchmark. Currently, Qwen3.6 Plus by Alibaba Cloud / Qwen Team leads with a score of 0.935. The average score across all models is 0.917.

What is the highest RefCOCO-avg score?

The highest RefCOCO-avg score is 0.935, achieved by Qwen3.6 Plus from Alibaba Cloud / Qwen Team.

How many models are evaluated on RefCOCO-avg?

7 models have been evaluated on the RefCOCO-avg benchmark, with 0 verified results and 7 self-reported results.

What categories does RefCOCO-avg cover?

RefCOCO-avg is categorized under spatial reasoning, grounding, and vision. The benchmark evaluates image models.

What is the best open-source model on RefCOCO-avg?

Qwen3.6-27B by Alibaba Cloud / Qwen Team is the top-ranked open-source model on RefCOCO-avg, with a score of 0.925 (rank #2).

Which model offers the best value on RefCOCO-avg?

Among models scoring within 10% of the leader, Qwen3.5-35B-A3B from Alibaba Cloud / Qwen Team is the cheapest, at $0.25 per million input tokens with a score of 0.892.

How recent are the RefCOCO-avg leaderboard results?

The RefCOCO-avg leaderboard was last updated in June 2026 and currently includes 7 evaluated models.

All benchmarks

RefCOCO-avg

RefCOCO-avg measures object grounding accuracy averaged across RefCOCO, RefCOCO+, and RefCOCOg benchmarks.

Qwen3.6 Plus from Alibaba Cloud / Qwen Team currently leads the RefCOCO-avg leaderboard with a score of 0.935 across 7 evaluated AI models.