Skip to content

Release history

marqo releases

Ecommerce Search and Discovery - marqo.ai

Back to tool Latest release

All releases

33 shown

2.26.0 Breaking risk
Notable features
  • Custom score rerankers in scoreModifiers via BM25 or vector closeness
  • lexicalOperand parameter for controlling lexical operators (or, and, weakAnd)
  • CONTAINS filter for token-level matching on lexical fields
View release on GitHub
2.25.2 Bug fix

Adds configurable connection recycling to mitigate connection imbalance in long-running deployments.

View release on GitHub
2.25.1 New feature
Notable features
  • center parameter for reproducible recency scores
  • applyToSubqueries to control recency in hybrid search
Full changelog

2.25.1

New Features

  • Add center and applyToSubqueries parameters to recency scoring (https://github.com/marqo-ai/marqo/pull/1376)
    • center — A fixed Unix epoch timestamp (seconds) to use as the reference point instead of now(), enabling reproducible recency scores across queries
    • applyToSubqueries — Control which hybrid subqueries ("tensor", "lexical", or both) receive recency boosting in RRF hybrid search
View release on GitHub
2.25.0 New feature
Notable features
  • Triton-based inference orchestrator
  • Model Management Container with lifecycle management
  • Centralized model registry in marqo-common
Full changelog

2.25.0

Changes

Triton-based Inference Architecture

Marqo's inference layer has been restructured from a monolithic design into three dedicated components:

  • Inference Orchestrator — A FastAPI service that coordinates inference requests. (https://github.com/marqo-ai/marqo/pull/1315)
  • Model Management Container — A FastAPI service for managing ML model lifecycles with Triton Inference Server, including model loading/unloading, health checks, and environment variable consistency. (https://github.com/marqo-ai/marqo/pull/1322)
  • Marqo API adaptations — The core Marqo API has been updated to work with the new Triton-backed components. (https://github.com/marqo-ai/marqo/pull/1317)

This architecture enables independent scaling and deployment of inference, model management, and search API layers.

Other Changes

  • Centralize model registry into a shared components/common package (marqo-common). (https://github.com/marqo-ai/marqo/pull/1356)
  • Fix model download auth handling to support public S3 buckets. (https://github.com/marqo-ai/marqo/pull/1352)
View release on GitHub
2.24.15 Bug fix

Minor fixes and improvements.

Full changelog

2.24.15

Bug Fixes and Minor Changes

  • Back-port update_index_settings API to 2.24 release branch to support modifying modelProperties of an existing index (https://github.com/marqo-ai/marqo/pull/1369)
View release on GitHub
2.24.13 Bug fix

Minor fixes and improvements.

Full changelog

2.24.13

Bug Fixes and Minor Changes

  • Use orjson in get_document(s) endpoints (https://github.com/marqo-ai/marqo/pull/1349)
  • Support picking the representative document within each collapsed group based on a numeric field sorting result (https://github.com/marqo-ai/marqo/pull/1350)
View release on GitHub
2.24.11 New feature
Notable features
  • weakAnd lexical retrieval support
  • Second-phase score modifiers
Full changelog

2.24.11

Bug Fixes and Minor Changes

  • Support weankAnd lexical retrieval and second-phase score modifiers (https://github.com/marqo-ai/marqo/pull/1344).
View release on GitHub
2.24.10 New feature
Notable features
  • IPv6 socket support with IPv4 default
Full changelog

2.24.10

Bug Fixes and Minor Changes

  • Add support for IPV6 sockets and set the default to IPV4 (https://github.com/marqo-ai/marqo/pull/1341).
View release on GitHub
2.24.8 New feature
Notable features
  • Recency-based relevance score boosting
Full changelog

2.24.8

Bug Fixes and Minor Changes

  • Support boosting the relevance score based on the recency of a document (#1330).
View release on GitHub
2.24.7 Bug fix

Fixed performance degradation when using collapsingFields with attributesToRetrieve.

View release on GitHub
2.24.5 Bug fix

Fixed _relevantCandidates changing unexpectedly with relevance cut-off and collapse field.

View release on GitHub
2.24.0 New feature
Notable features
  • Typeahead search with intelligent suggestions
  • Caching for base64-encoded image embeddings
View release on GitHub
2.23.2 Bug fix

Fixes totalHits count in collapsed searches to reflect unique values instead of total matching documents.

View release on GitHub
2.23.0 New feature
Notable features
  • Configurable stemming for text fields
  • Collapse fields for variant grouping
View release on GitHub
2.22.0 New feature
Notable features
  • Search result sorting with sortBy parameter
  • Personalization with context documents
  • Query logging for slow and failed queries
View release on GitHub
2.21.1 Bug fix

Bug fix that omits base64 image strings from search responses to reduce response size.

View release on GitHub
2.21.0 New feature
Notable features
  • Base64-encoded image search
  • Language support for lexical fields
View release on GitHub
2.19.3 Bug fix

Fixed bug affecting recall when combining searchable_attributes with required terms in lexical and hybrid queries.

View release on GitHub
2.19.1 Bug fix

Improved model warm-up logic to reduce memory usage by warming models on a single device.

View release on GitHub
2.20.0 New feature
Notable features
  • Inference cache (40% throughput, 25% latency improvement)
  • SigLIP2 model support
  • Approximate threshold parameter
View release on GitHub

Beta — feedback welcome: [email protected]