![](/catalog-art/retrievers-header.webp)

1.  [](/)
2.  [Retrievers](/retrievers)

# Retrievers

Retrievers combine the features your extractors produce into multi-stage search pipelines. Each stage below is one step you can chain: search, filter, rank, enrich, reduce.

For agents`Build a multi-stage search with create_retriever on the Mixpeek MCP server (https://mcp.mixpeek.com/mcp) (stages such as feature_search, attribute_filter and rerank), then query it with execute_retriever. Every MCP tool needs an API key; an agent can get one with POST https://api.mixpeek.com/v1/signup/agent.`Copy

Search stages...⌘K

[Search your own data](https://studio.mixpeek.com/namespaces/create?mode=managed&utm_source=homepage&utm_medium=retrievers_index&utm_campaign=mvs_onramp) [Retriever docs](https://mixpeek.com/docs/retrieval/stages/overview)

All stagesApplyReduceFilterSortEnrich

A retriever is the stable multi-stage contract an application binds to by name, so the embedding model, the index and the query plan underneath can all change without breaking a caller. That job is one layer of [the semantic layer for unstructured and multimodal data](/semantic-layer), the architecture that turns raw objects into named features and serves them through contracts an application can bind to.

## Filters

Reset

Stage type

*   Filter6
*   Sort5
*   Reduce9
*   Apply12
*   Enrich2

Filters

34 stages

*   ### [Feature Search](/retrievers/feature_search)
    
    Filter
    
    Search and filter documents by vector similarity using feature embeddings
    
    stage\_id feature\_search[Docs](https://mixpeek.com/docs/retrieval/stages/feature-search)View stage
    
*   ### [Attribute Filter](/retrievers/attribute_filter)
    
    Filter
    
    Filter documents by metadata attribute values using boolean logic
    
    stage\_id attribute\_filter[Docs](https://mixpeek.com/docs/retrieval/stages/attribute-filter)View stage
    
*   ### [Llm Filter](/retrievers/llm_filter)
    
    Filter
    
    Filter documents using LLM-based semantic evaluation
    
    stage\_id llm\_filter[Docs](https://mixpeek.com/docs/retrieval/stages/llm-filter)View stage
    
*   ### [Agent Search](/retrievers/agent_search)
    
    Filter
    
    LLM-driven multi-step retrieval with iterative reasoning and tool orchestration
    
    stage\_id agent\_search[Docs](https://mixpeek.com/docs/retrieval/stages/agent-search)View stage
    
*   ### [Query Expand](/retrievers/query_expand)
    
    Filter
    
    Generate query variations with LLM and fuse search results via RRF
    
    stage\_id query\_expand[Docs](https://mixpeek.com/docs/retrieval/stages/query-expand)View stage
    
*   ### [Score Threshold](/retrievers/score_threshold)
    
    Filter
    
    Drop results below an absolute score and return no results when nothing qualifies
    
    stage\_id score\_threshold[Docs](https://mixpeek.com/docs/retrieval/stages/score-threshold)View stage
    
*   ### [Sort Relevance](/retrievers/sort_relevance)
    
    Sort
    
    Sort documents by relevance score
    
    stage\_id sort\_relevance[Docs](https://mixpeek.com/docs/retrieval/stages/sort-relevance)View stage
    
*   ### [Sort Attribute](/retrievers/sort_attribute)
    
    Sort
    
    Sort documents by any metadata field value
    
    stage\_id sort\_attribute[Docs](https://mixpeek.com/docs/retrieval/stages/sort-attribute)View stage
    
*   ### [Mmr](/retrievers/mmr)
    
    Sort
    
    Reorder results using Maximal Marginal Relevance for diversity
    
    stage\_id mmr[Docs](https://mixpeek.com/docs/retrieval/stages/mmr)View stage
    
*   ### [Rerank](/retrievers/rerank)
    
    Sort
    
    Rerank documents using cross-encoder models for accurate relevance
    
    stage\_id rerank[Docs](https://mixpeek.com/docs/retrieval/stages/rerank)View stage
    
*   ### [Score Normalize](/retrievers/score_normalize)
    
    Sort
    
    Rescale document scores to a common range for consistent comparison
    
    stage\_id score\_normalize[Docs](https://mixpeek.com/docs/retrieval/stages/score-normalize)View stage
    
*   ### [Aggregate](/retrievers/aggregate)
    
    Reduce
    
    Compute aggregations (COUNT, SUM, AVG, etc.) on pipeline results
    
    stage\_id aggregate[Docs](https://mixpeek.com/docs/retrieval/stages/aggregate)View stage
    

Page 1 of 3

## How Retriever Pipelines Work

Combine stages to build sophisticated search and retrieval pipelines. Each stage type serves a specific purpose in the data flow.

![What is a retriever pipeline: a query (text, image, video, or audio) flows through five chained stage groups (Search (feature\_search, query\_expand, agent\_search), Filter (attribute\_filter, llm\_filter, deduplicate), Rank (rerank, mmr, sort\_relevance, score\_normalize), Enrich (llm\_enrich, taxonomy\_enrich, rag\_prepare, document\_enrich), and Reduce (aggregate, group\_by, cluster, sample, limit, summarize)) producing scored, enriched, shaped results from 34 available stages in under 100ms p95. Each stage is JSON: stage\_type plus a config with stage\_id and parameters. Retrievers are stored, versioned, callable as one endpoint, and exposed to LLM agents as MCP tools.](/diagrams/retriever-pipeline.webp)

The full anatomy: stages chain sequentially (search, filter, rank, enrich, reduce), each processing the output of the one before it. Define the pipeline once in JSON, then call it from any client, or let an agent call it as an MCP tool.

### 1\. Search

Retrieve documents from collections using semantic search, keyword search, or hybrid approaches.

### 2\. Filter

Remove documents that don't match criteria using attribute filters, score thresholds, or LLM evaluation.

### 3\. Rank

Reorder documents by relevance using sorting, cross-encoders, or LLM-based reranking.

### 4\. Generate

Create responses from retrieved documents using RAG, summarization, or custom LLM generation.

## Build a retriever from the docs

The stages described above, with request shapes, and what each one passes to the next.

[RetrieversCreating a named retriever and calling it by name from your app.](https://docs.mixpeek.com/retrieval/retrievers)[Multi-stage retrieval deep diveHow stages compose, and what each one passes to the next.](https://docs.mixpeek.com/retrieval/multi-stage-deep-dive)[Feature search stageThe search stage itself. API reference and worked examples.](https://docs.mixpeek.com/retrieval/stages/feature-search)[Rerank stageReordering candidates with a cross-encoder before you return them.](https://docs.mixpeek.com/retrieval/stages/rerank)

## All retriever stages

*   [Agent Search](/retrievers/agent_search)
*   [Agentic Enrich](/retrievers/agentic_enrich)
*   [Aggregate](/retrievers/aggregate)
*   [API Call](/retrievers/api_call)
*   [Attribute Filter](/retrievers/attribute_filter)
*   [Classify](/retrievers/classify)
*   [Cluster](/retrievers/cluster)
*   [Code Execution](/retrievers/code_execution)
*   [Cross Compare](/retrievers/cross_compare)
*   [Deduplicate](/retrievers/deduplicate)
*   [Document Enrich](/retrievers/document_enrich)
*   [External Web Search](/retrievers/external_web_search)
*   [Feature Search](/retrievers/feature_search)
*   [Group By](/retrievers/group_by)
*   [Json Transform](/retrievers/json_transform)
*   [Limit](/retrievers/limit)
*   [LLM Enrich](/retrievers/llm_enrich)
*   [LLM Filter](/retrievers/llm_filter)
*   [MMR](/retrievers/mmr)
*   [Moment Group](/retrievers/moment_group)
*   [Query Expand](/retrievers/query_expand)
*   [RAG Prepare](/retrievers/rag_prepare)
*   [Rerank](/retrievers/rerank)
*   [Sample](/retrievers/sample)
*   [Score Normalize](/retrievers/score_normalize)
*   [Score Threshold](/retrievers/score_threshold)
*   [Sort Attribute](/retrievers/sort_attribute)
*   [Sort Relevance](/retrievers/sort_relevance)
*   [SQL Lookup](/retrievers/sql_lookup)
*   [Summarize](/retrievers/summarize)
*   [Taxonomy Enrich](/retrievers/taxonomy_enrich)
*   [Temporal](/retrievers/temporal)
*   [Unwind](/retrievers/unwind)
*   [Web Scrape](/retrievers/web_scrape)