Benchmark
MIEB(lite) - Any2AnyRetrieval
Vision benchmarks measure how well models understand images. Scores are percentages.
- Models measured
- 4
- Scores recorded
- 4
- Sources
- 1
- IQ weight
- No weight published
Score distribution
- 29.2%
- Mean
- 27.6%
- Median
- 5.5%
- Std dev
- 23.6%
- 10th percentile
- 38.3%
- 90th percentile
- 23.6%
- Min
Leaderboard
| # | Model | Score |
|---|---|---|
| 1 | embeddinggemma-2Any2AnyRetrieval· MTEB Multimodal Boards data docs | 38.3% |
| 2 | Jina Embeddings v5 Omni SmallAny2AnyRetrieval· MTEB Multimodal Boards data docs | 27.6% |
| 3 | Jina Clip v1Any2AnyRetrieval· MTEB Multimodal Boards data docs | 27.3% |
| 4 | Jina Embeddings v5 Omni NanoAny2AnyRetrieval· MTEB Multimodal Boards data docs | 23.6% |