Overview

Introduction

The Mixedbread Reranking API provides access to powerful models designed to significantly improve the relevance of search results and the quality of context returned in retrieval pipelines. While initial retrieval methods efficiently fetch a broad set of potential candidates, reranking applies a more sophisticated, semantically-aware model to re-order this initial set. Our API simplifies the integration of this crucial refinement step into your applications.

Typical Workflow: Using Reranking

Key Features

• State-of-the-Art Models: Access powerful reranking models like mxbai-rerank-large-v2 featuring reinforcement learning training and multilingual capabilities.
• Improved Precision: Significantly boosts the relevance of top search results compared to initial retrieval alone.
• Text & Structured Data: Rerank not only text documents but also structured objects by specifying relevant text fields.
• Easy Integration: Simple API for seamless implementation into existing search workflows.
• Continuous Updates: API models benefit from regular updates and fine-tuning on fresh data.

Best Practices for Reranking

1. Diverse Initial Results: Make sure your first-stage retrieval casts a wide net. Reranking can't magically create relevant results if they're not in the initial set.

2. Experiment with Reranking Depth: Try reranking different numbers of initial results (e.g., top 10, top 50, top 100) to find the sweet spot for your use case.

3. Combine with Other Techniques: Use reranking alongside other relevance boosting methods like query expansion or faceted search for even better results.

4. Monitor and Iterate: Keep an eye on your search metrics and user feedback. Reranking might reveal insights about your content or user needs that you can use to further improve your system.

5. Consider Specialized Use Cases: Our reranking models excel beyond document search, including code snippets, SQL ranking, LLM tool selection, and e-commerce product ranking.

Last updated: June 11, 2025