Vai al contenuto
AI.info

AI Tools

Mixedbread Rerank

Mixedbread Rerank riordina i risultati iniziali della ricerca di Store con un modello di seconda fase per migliorarne la pertinenza.

Il sito di Mixedbread Rerank

In inglese

Key facts

Price
Free plan · paid plans from $20/mo Read on mixedbread.com
Free plan
Yes · Starter Read on mixedbread.com
Platforms
Web · API · CLI
API
Yes Read on mixedbread.com
Open source
Not stated on the maker’s pages
Latest version
mixedbread-ai/mxbai-rerank-v3.1-listwise
Last update
10 Nov 2026 Read on mixedbread.com
Prices checked
19 Nov 2026 Read on mixedbread.com

Mixedbread Rerank is a feature of Mixedbread Search. Developers enable it in Store search requests to rescore and reorder retrieved results, choose a supported reranking model, set a result limit, and include selected metadata.

It is available through Mixedbread’s API and SDKs. Reranking adds latency and usage charges: the pricing page lists semantic search with reranking at $7.50 per 1K queries. The feature does not itself create or index the search data.

Category
Developer Tools
Read from the maker’s pages
2026-11-18

Company

Mixedbread

Features

  • Enable reranking with a boolean option in Store search requests.
  • Choose a supported reranking model; the default is mixedbread-ai/mxbai-rerank-v3.1-listwise.
  • Set how many retrieved results to return after reranking.
  • Include all metadata or select specific metadata fields.
  • Use the Python, TypeScript, or cURL examples in the documentation.

Use cases

  • Improve the ordering of search results for complex queries.
  • Re-rank retrieved documents before passing them to an AI assistant.
  • Prioritize results using a supported reranking model.
  • Return selected metadata alongside reranked results.

Pricing

Free plan · paid plans from $20/mo

Prices checked
2026-11-19

Starter

free

  • 3 workspace users
  • 3 stores
  • 100 queries/min, 6,000 uploads/min
  • Community Slack support

Scale

$20/mo

  • $20 / month
  • $20 of credits included every month
  • Unlimited workspace users
  • 10,000 stores
  • 2,400 queries/min, 12,000 uploads/min
  • Automatic backups
  • Priority Slack support (same-day SLA)

Enterprise

price on request

  • Volume-based discounts
  • Unlimited workspace users
  • Unlimited stores
  • Custom rate limits & SLA
  • Automatic backups, point-in-time recovery
  • Dedicated support team

Capabilities

  • Choice of models — “Reranking Models : Use any supported reranking model.” source
  • API — “Mixedbread is an API for integrating fast, multimodal search into your applications, agents, and AI systems.” source
  • Official SDKs — “from mixedbread import Mixedbread” source
  • Runs models for you — “This uses our default reranking model ( mixedbread-ai/mxbai-rerank-v3.1-listwise ) with standard settings to improve your search results.” source
  • Search over your data — “Query your store using natural language.” source

Latest updates

  • Cheaper Ingestion and Storage

    Ingestion and storage rates decreased, with the new rates applying automatically from the current billing period.

  • Speaker Labels in Audio and Video

    Audio and video transcriptions now label speakers, and m4a, aac, and webm audio are supported.

  • New File Types

    Uploads now support JSON, JSON Lines, XML, XBRL filings, and Outlook .msg emails with filterable email metadata.

  • Better HTML Ingestion

    HTML files are parsed as documents and returned as clean Markdown; billing is based on extracted text.

  • Domain Filters for Web Search

    The mixedbread/web Store supports domain filters, and web results now include metadata.domain.

Where it works

Web · API · CLI

Alternatives to Mixedbread Rerank

  • Mixedbread Embed

    Mixedbread’s embedding models turn text into vectors for retrieval; its standalone embedding API documentation is now listed as legacy.

  • Cohere

    Cohere provides enterprise AI models and tools for text generation, search, document parsing, speech transcription, and workplace agents.

  • Qdrant

    Qdrant stores and searches vector data for semantic, similarity, keyword, and hybrid retrieval in AI applications.

  • Pinecone

    Pinecone is a managed vector database for semantic, keyword, and hybrid search in AI applications.

  • Zilliz Cloud

    Managed vector database for storing, indexing, and searching embeddings in AI applications.

  • Weaviate

    Weaviate is an open-source vector database for storing and searching data in AI applications, with self-hosted and managed cloud options.

More from Mixedbread

  • Mixedbread Embed

    Mixedbread’s embedding models turn text into vectors for retrieval; its standalone embedding API documentation is now listed as legacy.

Questions about Mixedbread Rerank

Is Mixedbread Rerank free?

Yes. Mixedbread Rerank has a free plan, called Starter. Paid plans start at $20/mo. Prices checked on 19 Nov 2026.

How much does Mixedbread Rerank cost?

The plans on the maker’s pricing page: Starter (free), Scale ($20/mo) and Enterprise (price on request). Prices checked on 19 Nov 2026.

Does Mixedbread Rerank have an API?

Yes. The maker’s pages state that Mixedbread Rerank has an API. Read on the maker’s pages on 18 Nov 2026.

Which platforms does Mixedbread Rerank run on?

Mixedbread Rerank runs on Web, API and CLI, as its maker’s pages state. Read on the maker’s pages on 18 Nov 2026.

Is Mixedbread Rerank still available?

Yes, as of 18 Nov 2026: its maker’s pages were read that day and carried no notice that Mixedbread Rerank has ended.

What changed recently in Mixedbread Rerank?

The latest entry in its maker’s changelog: 10 Nov 2026, “Cheaper Ingestion and Storage”. Before it: 20 Oct 2026, “Speaker Labels in Audio and Video”. Changelog checked on 16 Nov 2026.

Sources

Not confirmed

Open source
The current product page does not state a license for the default reranking model.
Official website