---
language:
- en
license: mit
pretty_name: Moorcheh Vector Search Benchmarks
task_categories:
- text-retrieval
tags:
- moorcheh
- benchmark
- vector-database
- ndcg
- latency
configs:
- config_name: default
  data_files:
  - split: ndcg10_vector_fp
    path: ndcg_vector_fp.csv

  - split: ndcg10_vector_fp_categorized
    path: ndcg_vector_fp_categorized.csv

  - split: latency_vector_fp
    path: latency_vector_fp.csv

  - split: ndcg10_vector_quantized
    path: ndcg_vector_quantized.csv

  - split: ndcg10_vector_quantized_categorized
    path: ndcg_vector_quantized_categorized.csv

  - split: latency_vector_quantized
    path: latency_vector_quantized.csv
---


## Dataset Coverage

The benchmarks evaluate performance across 14+ specialized datasets covering:
- **Legal & Regulatory**: ACORDAR, AILA2019-Case, AILA2019-Statutes, LeCaRDv2, LegalQuAD, REGIR-EU2UK, REGIR-UK2EU.
- **Financial**: ConvFinQA, FinanceBench, FinQA, FiQA, HC3Finance.
- **Medical & Clinical**: NFCorpus.
- **General/API**: Apple Documentation.

## Metrics
1. **NDCG@10**: Normalized Discounted Cumulative Gain at rank 10, measuring retrieval quality.
2. **Latency (ms)**: Mean search latency measured on the server-side and end-to-end.

## Benchmark Configurations

The data is organized into four primary splits, which you can switch between using the "Viewer" tab above:

1. **NDCG@10 - Floating Point (`ndcg_vector_fp`)**: Evaluation of retrieval accuracy using standard 32-bit floating-point vectors.

![NDCG@10 Performance of Floating-Point Vector Embeddings on MAIR Datasets](https://cdn-uploads.huggingface.co/production/uploads/6818ea5217c253b3f852468b/bT16HC6sjfuL5DTjZLg7h.png)

2. **NDCG@10 – Floating Point (Categorized) (`ndcg_vector_fp_categorized`)**: Evaluation of retrieval accuracy using standard 32-bit floating-point vectors, reported per dataset category.

![NDCG@10 Performance of Floating-Point Vector Embeddings - Categorized](https://cdn-uploads.huggingface.co/production/uploads/6818ea5217c253b3f852468b/VRGrtCrxAg1V0YsfVqr8r.png)

3. **Latency - Floating Point (`latency_vector_fp`)**: Search performance (ms) for floating-point vector retrieval. 
4. **NDCG@10 - Quantized (`ndcg_vector_quantized`)**: Evaluation of retrieval accuracy using optimized quantized/binary embedding paths.

![NDCG@10 Performance of Quantized Vector Embeddings on MAIR Datasets](https://cdn-uploads.huggingface.co/production/uploads/6818ea5217c253b3f852468b/XkCI0Xj8Q3PAXLFoiVTrh.png)

5. **NDCG@10 – Quantized (Categorized) (`ndcg_vector_quantized_categorized`)**: Evaluation of retrieval accuracy using optimized quantized or binary vector representations, reported per dataset category.

![NDCG@10 Performance of Quantized Embeddings - Categorized](https://cdn-uploads.huggingface.co/production/uploads/6818ea5217c253b3f852468b/9NVbK2nwCICDNttT3lyJR.png)

6. **Latency - Quantized (`latency_vector_quantized`)**: Search performance (ms) for optimized quantized vector retrieval.

## Tested Providers
Results include comparisons across:
- **Moorcheh**
- **Elasticsearch**
- **Pinecone** (with Cohere Rerank)
- **PGVector** (PostgreSQL)
- **Qdrant**

## How to use in Python

You can load these results directly into a Pandas DataFrame using the Hugging Face `datasets` library:

```python
from datasets import load_dataset

# Load the Latency results for Floating Point vectors
dataset = load_dataset("moorcheh/Benchmarks", split="latency_vector_fp")
df = dataset.to_pandas()

print(df.head())
```

## Citation

If you use this dataset, please cite:

- [arXiv:2601.11557](https://arxiv.org/abs/2601.11557)