reasoning benchmark
ERQA is a reasoning benchmark with 400 multiple-choice visual questions covering spatial reasoning, trajectory reasoning, action reasoning, state estimation, and multi-view reasoning.
Updated Sep 5, 2026
Higher score ranks better on this benchmark.
Rank | Model | Score | Percentile | Participants | Evidence | Evaluated |
|---|
| Rank01 | ModelAC | Score77.80% | Percentile100.00% | Participants26 | EvidenceC | Evaluated |
| Rank02 | ModelAC | Score72.30% | Percentile96.00% | Participants26 | EvidenceC | Evaluated |
| Rank03 | ModelAC | Score72.30% | Percentile92.00% | Participants26 | EvidenceC | Evaluated |
| Rank04 | ModelBY | Score72.00% | Percentile88.00% | Participants26 | EvidenceC | Evaluated |
| Rank05 | ModelBY | Score71.30% | Percentile84.00% | Participants26 | EvidenceC | Evaluated |
| Rank06 | ModelAC | Score69.80% | Percentile80.00% | Participants26 | EvidenceC | Evaluated |
| Rank07 | ModelOP | Score65.70% | Percentile76.00% | Participants26 | EvidenceC | Evaluated |
| Rank08 | ModelAC | Score65.70% | Percentile72.00% | Participants26 | EvidenceC | Evaluated |
| Rank09 | ModelAC | Score65.50% | Percentile68.00% | Participants26 | EvidenceC | Evaluated |
| Rank10 | ModelAC | Score64.80% | Percentile64.00% | Participants26 | EvidenceC | Evaluated |
| Rank11 | ModelME | Score64.70% | Percentile60.00% | Participants26 | EvidenceC | Evaluated |
| Rank12 | ModelOP | Score64.00% | Percentile56.00% | Participants26 | EvidenceC | Evaluated |
| Rank13 | ModelAC | Score62.50% | Percentile52.00% | Participants26 | EvidenceC | Evaluated |
| Rank14 | ModelAC | Score62.00% | Percentile48.00% | Participants26 | EvidenceC | Evaluated |
| Rank15 | ModelAC | Score60.50% | Percentile44.00% | Participants26 | EvidenceC | Evaluated |
| Rank16 | ModelAC | Score52.50% | Percentile40.00% | Participants26 | EvidenceC | Evaluated |
| Rank17 | ModelAC | Score52.30% | Percentile36.00% | Participants26 | EvidenceC | Evaluated |
| Rank18 | ModelAC | Score51.30% | Percentile32.00% | Participants26 | EvidenceC | Evaluated |
| Rank19 | ModelAC | Score48.80% | Percentile28.00% | Participants26 | EvidenceC | Evaluated |
| Rank20 | ModelAC | Score47.30% | Percentile24.00% | Participants26 | EvidenceC | Evaluated |
| Rank21 | ModelAC | Score46.80% | Percentile20.00% | Participants26 | EvidenceC | Evaluated |
| Rank22 | ModelAC | Score45.80% | Percentile16.00% | Participants26 | EvidenceC | Evaluated |
| Rank23 | ModelAC | Score45.30% | Percentile12.00% | Participants26 | EvidenceC | Evaluated |
| Rank24 | ModelAC | Score43.00% | Percentile8.00% | Participants26 | EvidenceC | Evaluated |
| Rank25 | ModelAC | Score41.30% | Percentile4.00% | Participants26 | EvidenceC | Evaluated |
| Rank26 | ModelOP | Score35.20% | Percentile0.00% | Participants26 | EvidenceC | Evaluated |
The leading models and scores on this benchmark.
A closer view of the leading scores on this benchmark.
The first five results on this benchmark, with official price and output speed added where the model identity can be matched.
What ERQA measures and how its scores work.
ERQA is an Embodied Reasoning Question Answering benchmark for evaluating AI capabilities in physical world interactions.
ERQA measures performance on spatial reasoning, trajectory reasoning, action reasoning, state estimation, and multi-view reasoning using a Score reported as a ratio.
Scores are shown in ratio. This benchmark is not independently verified and has an evidence level of C.
LLM Stats. Benchmark scores retain their original unit. Overall score eligibility is shown separately.
Common questions about ERQA.
Qwen3.8 Max is currently ranked first with 77.80%.
The current leaders are Qwen3.8 Max (77.80%), Qwen3.8-Flash-Next (72.30%), and Qwen3.8 Flash (72.30%).
Qwen3.8 Flash has the lowest matched official input price at $0.15 input / $0.47 output per 1M tokens.
The fastest matched records are GPT-4o (132.00 tok/s via OpenAI), GPT-5 (100.00 tok/s via OpenAI), and Qwen3.8 Max (60.87 tok/s via DeepInfra).
No. This benchmark measures one defined capability or task. The overall LLMBoard score uses a separate aggregation across eligible benchmark evidence.
ERQA measures performance on spatial reasoning, trajectory reasoning, action reasoning, state estimation, and multi-view reasoning using a Score reported as a ratio.
Yes. Higher values rank better for this benchmark.
26 model results are currently shown.
No. This benchmark is shown for reference but does not contribute to the overall score.
Ranking basisThis erqa AI model leaderboard uses descending score in the benchmark original unit. The leaderboard ranking keeps matched price and speed data separate from benchmark evidence.
Selection summary
Qwen3.8 Max currently leads ERQA with 77.80%. 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.