spatial reasoning benchmark
RefCOCO-avg Leaderboard
RefCOCO-avg measures object grounding accuracy averaged across RefCOCO, RefCOCO+, and RefCOCOg benchmarks.
Updated Aug 17, 2026
Models9
Model coverage9
MetricScore
EvidenceB
RefCOCO-avg Ranking
Higher score ranks better on this benchmark.
Sort by
Rank | Model | Score | Percentile | Participants | Evidence | Evaluated |
|---|
| Rank01 | ModelAC | Score0.935 points | Percentile100.0% | Participants9 | EvidenceC | Evaluated |
| Rank02 | ModelAC | Score0.925 points | Percentile87.5% | Participants9 | EvidenceC | Evaluated |
| Rank03 | ModelAC | Score0.924 points | Percentile75.0% | Participants9 | EvidenceC | Evaluated |
| Rank04 | ModelAC | Score0.92 points | Percentile62.5% | Participants9 | EvidenceC | Evaluated |
| Rank05 | ModelAC | Score0.913 points | Percentile50.0% | Participants9 | EvidenceC | Evaluated |
| Rank06 | ModelAC | Score0.909 points | Percentile37.5% | Participants9 | EvidenceC | Evaluated |
| Rank07 | ModelAC | Score0.892 points | Percentile25.0% | Participants9 | EvidenceC | Evaluated |
| Rank08 | ModelLA | Score0.879 points | Percentile12.5% | Participants9 | EvidenceC | Evaluated |
| Rank09 | ModelCO | Score0.732 points | Percentile0.0% | Participants9 | EvidenceC | Evaluated |
RefCOCO-avg Highlights
The leading models and scores on this benchmark.
RefCOCO-avg Score Distribution
A closer view of the leading scores on this benchmark.
The Top AI Models for RefCOCO-avg
The first five results on this benchmark, with official price and output speed added where the model identity can be matched.
Ranking basisThis refcoco-avg AI model leaderboard uses descending score in the benchmark's original unit. The leaderboard ranking keeps matched price and speed data separate from benchmark evidence.
- 01ACQwen3.6 PlusAlibaba Cloud / Qwen Team
- Score
- 0.935 points
- Price
- $0.50 input / $3.0 output per 1M tokens
- Speed
- Up to 16 tok/s via Together
Strengths
- Ranks #1 of 9 compared models
- 100th percentile on this benchmark
- C evidence result
Considerations
- This result measures RefCOCO-avg, not total model capability
- 02ACQwen3.6-27BAlibaba Cloud / Qwen Team
- Score
- 0.925 points
- Price
- $0.60 input / $3.6 output per 1M tokens
- Speed
- Up to 6.1 tok/s via Novita
Strengths
- Ranks #2 of 9 compared models
- 88th percentile on this benchmark
- C evidence result
Considerations
- This result measures RefCOCO-avg, not total model capability
- 03ACQwen3 VL 235B A22B ThinkingAlibaba Cloud / Qwen Team
- Score
- 0.924 points
Strengths
- Ranks #3 of 9 compared models
- 75th percentile on this benchmark
- C evidence result
Considerations
- This result measures RefCOCO-avg, not total model capability
- 04ACQwen3.6-35B-A3BAlibaba Cloud / Qwen Team
- Score
- 0.92 points
- Price
- $0.25 input / $1.5 output per 1M tokens
Strengths
- Ranks #4 of 9 compared models
- 63th percentile on this benchmark
- C evidence result
Considerations
- This result measures RefCOCO-avg, not total model capability
- 05ACQwen3.5-122B-A10BAlibaba Cloud / Qwen Team
- Score
- 0.913 points
- Price
- $0.40 input / $3.2 output per 1M tokens
Strengths
- Ranks #5 of 9 compared models
- 50th percentile on this benchmark
- C evidence result
Considerations
- This result measures RefCOCO-avg, not total model capability
Selection summary
Best AI Models for RefCOCO-avg
Qwen3.6 Plus currently leads RefCOCO-avg with 0.935 points. It is the top model on this specific benchmark, while the best LLM for the broader task should also be checked against other benchmarks, price and runtime.
Use this leaderboard with the supporting benchmark results and coverage details above. A leaderboard position summarizes the selected ranking signal; it does not replace workload-specific testing.
What is RefCOCO-avg?
What RefCOCO-avg measures and how its scores work.
RefCOCO-avg measures object grounding accuracy averaged across RefCOCO, RefCOCO+, and RefCOCOg benchmarks.
Scores are shown in points. This benchmark is not independently verified and has an evidence level of B.
- Family
- RefCOCO-avg
- Modality
- image
- Primary category
- spatial reasoning
- Score direction
- higher
- LLMBoard eligible
- Yes
- Evaluation key
- refcoco-avg|llm-stats-current
Benchmark scores retain their original unit. Overall score eligibility is shown separately.
FAQ
Common questions about RefCOCO-avg.
Which model scores highest on RefCOCO-avg?
Qwen3.6 Plus is currently ranked first with 0.935 points.
What does RefCOCO-avg measure?
RefCOCO-avg measures object grounding accuracy averaged across RefCOCO, RefCOCO+, and RefCOCOg benchmarks.
Is a higher score better?
Yes. Higher values rank better for this benchmark.
How many models are compared?
9 model results are currently shown.
Does this benchmark affect the overall score?
Yes. This benchmark can contribute to the current LLMBoard capability score.