vision benchmark
ODinW Leaderboard
ODinW is a benchmark for evaluating object detection models' task-level transfer ability across diverse real-world datasets.
Updated Sep 7, 2026
Models16
Model coverage16
MetricScore
EvidenceB
ODinW Ranking
Higher score ranks better on this benchmark.
Sort by
Rank | Model | Score | Percentile | Participants | Evidence | Evaluated |
|---|
| Rank01 | ModelAC | Score51.80% | Percentile100.00% | Participants16 | EvidenceC | Evaluated |
| Rank02 | ModelAC | Score51.10% | Percentile93.33% | Participants16 | EvidenceC | Evaluated |
| Rank03 | ModelAC | Score50.80% | Percentile86.67% | Participants16 | EvidenceC | Evaluated |
| Rank04 | ModelAC | Score48.60% | Percentile80.00% | Participants16 | EvidenceC | Evaluated |
| Rank05 | ModelAC | Score48.20% | Percentile73.33% | Participants16 | EvidenceC | Evaluated |
| Rank06 | ModelAC | Score47.50% | Percentile66.67% | Participants16 | EvidenceC | Evaluated |
| Rank07 | ModelAC | Score46.60% | Percentile60.00% | Participants16 | EvidenceC | Evaluated |
| Rank08 | ModelAC | Score44.70% | Percentile53.33% | Participants16 | EvidenceC | Evaluated |
| Rank09 | ModelAC | Score44.50% | Percentile46.67% | Participants16 | EvidenceC | Evaluated |
| Rank10 | ModelAC | Score43.20% | Percentile40.00% | Participants16 | EvidenceC | Evaluated |
| Rank11 | ModelAC | Score42.60% | Percentile33.33% | Participants16 | EvidenceC | Evaluated |
| Rank12 | ModelAC | Score42.40% | Percentile26.67% | Participants16 | EvidenceC | Evaluated |
| Rank13 | ModelAC | Score42.30% | Percentile20.00% | Participants16 | EvidenceC | Evaluated |
| Rank14 | ModelAC | Score41.10% | Percentile13.33% | Participants16 | EvidenceC | Evaluated |
| Rank15 | ModelAC | Score39.80% | Percentile6.67% | Participants16 | EvidenceC | Evaluated |
| Rank16 | ModelAC | Score39.40% | Percentile0.00% | Participants16 | EvidenceC | Evaluated |
ODinW Highlights
The leading models and scores on this benchmark.
ODinW Score Distribution
A closer view of the leading scores on this benchmark.
The Top AI Models for ODinW
The first five results on this benchmark, with official price and output speed added where the model identity can be matched.
What is ODinW?
What ODinW measures and how its scores work.
ODinW, or Object Detection in the Wild, is a vision benchmark for object detection models.
It measures prediction accuracy and adaptation efficiency, reported as a Score ratio.
Scores are shown in ratio. This benchmark is not independently verified and has an evidence level of B.
- Family
- ODinW
- Modality
- image
- Primary category
- vision
- Score direction
- higher
- LLMBoard eligible
- No
- Evaluation key
- overall
LLM Stats. Benchmark scores retain their original unit. Overall score eligibility is shown separately.
FAQ
Common questions about ODinW.
Which model scores highest on ODinW?
Qwen3.6 Plus is currently ranked first with 51.80%.
What are the top three models on ODinW?
The current leaders are Qwen3.6 Plus (51.80%), Qwen3.7-Plus (51.10%), and Qwen3.6-35B-A3B (50.80%).
Which ODinW model has the lowest official input price?
Qwen2.5-Omni-7B has the lowest matched official input price at $0.10 input / $0.40 output per 1M tokens.
Which models are fastest among ODinW results?
The fastest matched records are Qwen3 VL 4B Thinking (8.77 tok/s via DeepInfra).
Does the highest score result prove overall model quality?
No. This benchmark measures one defined capability or task. The overall LLMBoard score uses a separate aggregation across eligible benchmark evidence.