The git binary must now be present at build time to populate the commitSha1 field of the /version field.

This change was made for build performance reasons.

by @dureuill in https://github.com/meilisearch/meilisearch/pull/6030

The new settings indexer scales better, supports near-instant cancellations and displays the progress of the indexing operation.

Previously, the new settings indexer was enabled only if the only changes in a settings batch were to embedder settings. In Meilisearch v1.29.0, the new settings indexer will be enabled if the change is any combination of:

• searchableAttributes
• exactAttributes
• proximityPrecision
• embedders (as before)

Any other change to settings appearing in a batch will cause the batch to use the legacy settings indexer. Additionally, the new settings indexer is currently disabled by default for Cloud users.

OSS users who would like to disable the new settings indexer should pass the MEILI_EXPERIMENTAL_NO_EDITION_2024_FOR_SETTINGS environment variable to true.

by @Kerollmops in https://github.com/meilisearch/meilisearch/pull/5983

Meilisearch v1.21.0 introduced a new vector store backend providing better performance and relevancy.

Starting with v1.29.0, any newly created index will default to the new backend. Existing indexes will be left unchanged.

It is still possible to explicitly choose the vector store backend, please refer to the relevant experimental feature discussion page.

by @Kerollmops in https://github.com/meilisearch/meilisearch/pull/6004

You can now select models with the XLM Roberta architecture when generating embeddings locally on CPU or GPU with the huggingFace embedder.

by @qdequele in https://github.com/meilisearch/meilisearch/pull/6018

• Fix rare internal error in the experimental vector store by @Kerollmops in https://github.com/meilisearch/meilisearch/pull/6032

Last week was a Quality of Life week, and while we still had improvements in the pipe, the bulk of our work was dedicated to maintenance tasks.

Most notably, the CI is now faster, going from over one hour to less than 30 minutes, and also more reliable as it automatically tests the dumpless upgrade.

• New workload to run declarative tests. This is, in particular, very useful to automatically test the dumpless upgrade. By @Mubelotix in https://github.com/meilisearch/meilisearch/pull/5861
• The code that applies the dumpless upgrade migrations was rewritten to make it easier to maintain. By @dureuill in https://github.com/meilisearch/meilisearch/pull/6029
• We no longer run macOS and Windows tests at PR time. These CI jobs were very slow and given their limited impact they will only be run on the nightly builds. Ny @Kerollmops in https://github.com/meilisearch/meilisearch/pull/6021
• Send notifications for Kubernetes integration when releasing by @curquiza in https://github.com/meilisearch/meilisearch/pull/6023
• Introduce xtask sub-command to generate prototype tag names with the expected nomenclature by @Kerollmops in https://github.com/meilisearch/meilisearch/pull/6022
• Fix misc. CI issues by @dureuill in https://github.com/meilisearch/meilisearch/pull/6026

Full Changelog: https://github.com/meilisearch/meilisearch/compare/v1.28.2...v1.29.0

This release fixes a bug affecting the Prometheus metrics route in versions 1.28.0 and 1.28.1, specifically when the instance has too many tasks. The issue is visible as high memory usage and could cause the instance to be OOM-killed. If you are using the /metrics route, we recommend deleting all succeeded or failed functions in the index using the dedicated route or upgrading to at least v1.28.2.

• Limit the number of retrieved tasks to one in the metrics by @Kerollmops in https://github.com/meilisearch/meilisearch/pull/6024

Full Changelog: https://github.com/meilisearch/meilisearch/compare/v1.28.1...v1.28.2

This release features two fixes: one to ensure that we correctly upload the Linux-amd64 binaries for the Community and Enterprise editions, and one contributed by an external developer to ensure that we still return documents that don't contain the sortable attribute after those that do.

• Fix sort on /documents endpoint when field has no values by @EclipseAditya in https://github.com/meilisearch/meilisearch/pull/6016
• Fix release CI after we introduced the enterprise edition by @paulden in https://github.com/meilisearch/meilisearch/pull/6020

• @EclipseAditya made their first contribution in https://github.com/meilisearch/meilisearch/pull/6016

Full Changelog: https://github.com/meilisearch/meilisearch/compare/v1.28.0...v1.28.1

This release introduces improvements to language support and separates the community and enterprise binary editions. We now offer binaries under the BUSL-1.1 license, identified by the "enterprise" term in their names, in addition to our MIT-licensed binaries, which retain their original names. Docker images for the enterprise edition are available in the getmeili/meilisearch-enterprise repository.

• Separation of EE and CE. CE remains the default binary, and the name does not change by Louis on #6011

• Charabia v0.9.9: introduce a better word segmentation for Thai, Khmer, and German languages by @ManyTheFish in #6007
• Expose batch progress traces on the metrics route to improve the indexing debugging experience by @Kerollmops in #5956

• Remove version from the name of the test job in CI by @paulden in #6012
• Upgrade most of the dependencies by @Kerollmops in #6002
• Build x86 and ARM images on GitHub-hosted runners by @paulden in #6003
• Fix SDKs tests to use the enterprise edition and continue testing the sharding feature by @curquiza in #6013

• @paulden made their first contribution in #6003

Full Changelog: https://github.com/meilisearch/meilisearch/compare/v1.27.0...v1.28.0

This release improves the batch size for better performance. It also fixes bugs with the embedders that could skip some documents during generation and resolves an issue with the document route that displayed the same documents on multiple pages. It improves the quality of error messages when uploading snapshots to S3, which helps with debugging.

• Update macOS platform version in the CI by @Kerollmops in https://github.com/meilisearch/meilisearch/pull/6001 Meilisearch MacOS binaries now generated with MacOS Sonoma (macos-14)

• Show errors in the task queue to improve debugging of S3 snapshot uploads by @Kerollmops in https://github.com/meilisearch/meilisearch/pull/5994
• Batched tasks total size now defaults to half of the max indexing memory to improve task ingestion by @Kerollmops in https://github.com/meilisearch/meilisearch/pull/5990 & https://github.com/meilisearch/meilisearch/pull/6005

• Fix issue that could cause Meilisearch to skip some documents when performing embedding operations by @dureuill in https://github.com/meilisearch/meilisearch/pull/5995
  • Every available_parallelismth document in a batch was ignored for the purpose of embedding when using a Hugging Face embedder https://github.com/meilisearch/meilisearch/issues/5976
  • Every 40th document in a batch was ignored for the purpose of embedding when using a REST embedder with only one embedding per request
  • To verify if documents in your database have been affected:
    1. enable the multimodal exp. feature
    2. search or fetch with filter: NOT _vectors EXISTS to find documents without vectors.
• Fix /documents/fetch bug that could cause duplicated search results when paginating sorted documents by @ManyTheFish in https://github.com/meilisearch/meilisearch/pull/5999

Full Changelog: https://github.com/meilisearch/meilisearch/compare/v1.26.0...v1.27.0

• To make it easier to keep track of which documents were processed by Meilisearch, it is now possible to attach an arbitrary string to all routes that create document-related tasks.
• Tasks created with this custom metadata will display the passed metadata when accessed by the tasks route or sent in webhooks.
• To use this feature, add the customMetadata query parameter to any supported route:

POST /indexes/{indexUid}/documents?customMetadata=my-metadata-for-the-task

• Note that, as usual for query parameters, the value of the parameter must be URL-encoded.
• List of supported routes:

POST /indexes/{indexUid}/documents
PUT /indexes/{indexUid}/documents
DELETE /indexes/{indexUid}/documents/{documentId}
POST /indexes/{indexUid}/documents/delete-batch
POST /indexes/{indexUid}/documents/delete
POST /indexes/{indexUid}/documents/edit
DELETE /indexes/{indexUid}/documents

• Sample output of GET /tasks for tasks with metadata:

{
  "results": [
    {
      "uid": 37,
      "batchUid": 37,
      "indexUid": "mieli",
      "status": "succeeded",
      "type": "documentDeletion",
      "canceledBy": null,
      "details": {
        "deletedDocuments": 31944
      },
      "error": null,
      "duration": "PT0.511099S",
      "enqueuedAt": "2025-11-06T16:33:37.816237Z",
      "startedAt": "2025-11-06T16:33:37.821591Z",
      "finishedAt": "2025-11-06T16:33:38.33269Z",
      "customMetadata": "removeall"
    },
    {
      "uid": 36,
      "batchUid": 36,
      "indexUid": "movies",
      "status": "succeeded",
      "type": "documentAdditionOrUpdate",
      "canceledBy": null,
      "details": {
        "receivedDocuments": 31968,
        "indexedDocuments": 31944
      },
      "error": null,
      "duration": "PT3.192271S",
      "enqueuedAt": "2025-10-30T10:31:12.896073Z",
      "startedAt": "2025-10-30T10:31:12.911905Z",
      "finishedAt": "2025-10-30T10:31:16.104176Z",
      "customMetadata": "foo"
    }
  ],
  "total": 38,
  "limit": 2,
  "from": 36,
  "next": 35
}

by @dureuill in https://github.com/meilisearch/meilisearch/pull/5963

You can now select models with the modernBERT architecture when generating embeddings locally on CPU or GPU with the huggingFace embedder.

This unlocks for instance Ruri v3 and other models

by @hayatosc in https://github.com/meilisearch/meilisearch/pull/5980

You can now decide to ignore some embedder-related errors. Either:

1. Errors related to a document template not rendering properly
2. Errors related to an embedding request to an embedder failing (this includes missing vectors in userProvided embedders)
3. Or both kinds of errors.

When errors are ignored, the corresponding documents will not have embeddings, but the associated batch of tasks will not be marked as failed.

Of course, ignoring errors means that it is harder to notice an issue with embedders, so use this feature parsimoniously.

To enable the feature:

• Customers of the Cloud, please ask the support.
• OSS users, please use the MEILI_EXPERIMENTAL_CONFIG_EMBEDDER_FAILURE_MODES and set it to a comma-separated list of errors to ignore, with the possible values:
  1. ignore_document_template_failures to ignore document template failures
  2. ignore_embedder_failures to ignore embedder failures
  • For example: ignore_document_template_failures,ignore_embedder_failures ignores both kinds of failures

by @dureuill in https://github.com/meilisearch/meilisearch/pull/5984

You can now control the duration before a REST embedder request times out.

• Customers of the Cloud, please ask the support.
• OSS users, please use the MEILI_EXPERIMENTAL_REST_EMBEDDER_TIMEOUT_SECONDS, which must be a positive integer.

by @dureuill in https://github.com/meilisearch/meilisearch/pull/5984

• Remove unused dependency allocator-api2 by @xuhongxu96 in https://github.com/meilisearch/meilisearch/pull/5969

Many thanks to our new contributors @hayatosc and @xuhongxu96 ❤️

Add the ability to dynamically rerank the search results based on Cohere using a personalized prompt. For more information on how to set it up, see the dedicated experimental feature discussion.

Add the ability to upload snapshots directly to S3. It has many advantages, such as being able to stream the entire process and effectively utilizing multipart technology to send chunks of data in parallel. For more information on how to use it, see the dedicated experimental feature discussion.

The value of the Authorization header is now redacted when getting webhook, getting webhooks, or in the object returned when posting a new webhook or deleting a webhook.

• Impact on Cloud: Similarly to embedder API, the UI should make sure to never send the Authorization header back after it has been redacted
• Technically a breaking change, as users could previously get the key value back when getting the webhook, and that will no longer be possible

This bug was causing crashes in the recent indexer optimizations. If you deactivated these optimizations by setting the following environment variables:

MEILI_EXPERIMENTAL_NO_EDITION_2024_FOR_PREFIX_POST_PROCESSING=true
MEILI_EXPERIMENTAL_NO_EDITION_2024_FOR_FACET_POST_PROCESSING=true

You can now safely reactivate them without experiencing memory leaks.

German word segmentation relies on a word dictionary to segment words, but if a word is not part of the dictionary, it is cut into bigrams. The segmenter will now skip segmenting unknown words:

Source: "Feuchteschutz"
Before: ["Fe" "uc" "ht" "es, "ch", "utz"]
After:  ["Feuchteschutz"]

Source: "insgesamt"
Before: ["in" "sg" "es" "amt"]
After:  ["insgesamt"]

If you have a Meilisearch database containing German words, you must reindex your data manually.

It’s very common for Chinese, numbers, and English to appear together in the same sentence. We now ensure that numbers and English are not segmented differently between segmenters:

Source: "我从2025年开始学习Rust语言。"
Before: ["我", "从", "2", "0", "2", "5", "年", "开始", "学习", "R", "u", "s", "t", "语言", "。"]
After:  ["我", "从", "2025", "年", "开始", "学习", "Rust", "语言", "。"]

If you have a Meilisearch database containing Chinese words, you must reindex your data manually.

• Add Flickr example to README by @Kerollmops in https://github.com/meilisearch/meilisearch/pull/5961

Thanks to @JinheLin, @dureuill, @Kerollmops, and @ManyTheFish for their contribution! 🎉

This release features some improvements with the interaction of the vector store and the searchCutoffMs when using the vectorStore: experimental. It also introduces the metadata header Meili-Include-Metadata on the search request that adds a metadata field to the response. These metadatas contains one uid by query and a reminder of the indexUid and its primary key. We also introduced minor bug fixes around the compaction to improve the interaction with task cancellation.

• Improve the vector store search cutoff by @dureuill in https://github.com/meilisearch/meilisearch/pull/5945
• Improve compaction behaviors by @Kerollmops in https://github.com/meilisearch/meilisearch/pull/5946
• Search metadata by @ManyTheFish in https://github.com/meilisearch/meilisearch/pull/5926

• Adapt the standards of prototypes by @curquiza in https://github.com/meilisearch/meilisearch/pull/5942
• Bump Dockerfile alpine version to 3.22 by @PedroTroller in https://github.com/meilisearch/meilisearch/pull/5866

• @PedroTroller made their first contribution in https://github.com/meilisearch/meilisearch/pull/5866

Full Changelog: https://github.com/meilisearch/meilisearch/compare/v1.23.0...v1.24.0

This release introduces a new compact route on the index routes, which appends a new compaction task to the queue. Meilisearch uses an LMDB environment by index, and indexes start to fragment after some time. We have noticed that the indexes generally have 30% fragmentation. By defragmenting the environment, we've seen large (2-4x) speed-ups in terms of search and indexation. This is primarily due to the reordering of the LMDB internal pages and the removal of scattered free pages throughout the file, thereby relocating the content to the beginning.

We also worked on parallelizing the post-processing of facets. We noticed that a lot of time was spent iterating over the prefixes of the index in a single-threaded loop. We redesigned this part of the indexation to make it multi-threaded. We have seen a 4x and 6x improvement in terms of time spent on this operation.

• Introduce a task to compact an index by @Kerollmops in https://github.com/meilisearch/meilisearch/pull/5929
• Parallelize bulk facets & word prefix fid/position docids by @Kerollmops in https://github.com/meilisearch/meilisearch/pull/5307
• Change Java version in SDK CI by @curquiza in https://github.com/meilisearch/meilisearch/pull/5910
• Minor improvement in OpenAPI CI by @curquiza in https://github.com/meilisearch/meilisearch/pull/5834
• Add request uid to search routes by @ManyTheFish in https://github.com/meilisearch/meilisearch/pull/5863

• Fix ranking score bug when sort is present by @ManyTheFish in https://github.com/meilisearch/meilisearch/pull/5933
• Synonym performance fix by @ManyTheFish in https://github.com/meilisearch/meilisearch/pull/5930
• Update README.md to fix newsletter link by @EazyAl in https://github.com/meilisearch/meilisearch/pull/5911
• Try to fix GH license detection again by @dureuill in https://github.com/meilisearch/meilisearch/pull/5938

• Remove release-drafter and encourage usage of GitHub-generated notes by @curquiza in https://github.com/meilisearch/meilisearch/pull/5935
• Show Dependabot dependency upgrade in the changelog by @curquiza in https://github.com/meilisearch/meilisearch/pull/5900
• Bump actions/setup-go from 5 to 6 by @dependabot[bot] in https://github.com/meilisearch/meilisearch/pull/5912
• Bump actions/setup-dotnet from 4 to 5 by @dependabot[bot] in https://github.com/meilisearch/meilisearch/pull/5914
• Bump actions/setup-node from 4 to 5 by @dependabot[bot] in https://github.com/meilisearch/meilisearch/pull/5915
• Bump sigstore/cosign-installer from 3.9.2 to 3.10.0 by @dependabot[bot] in https://github.com/meilisearch/meilisearch/pull/5916
• Bump actions/setup-python from 5 to 6 by @dependabot[bot] in https://github.com/meilisearch/meilisearch/pull/5913

• @EazyAl made their first contribution in https://github.com/meilisearch/meilisearch/pull/5911

Full Changelog: https://github.com/meilisearch/meilisearch/compare/v1.22.1...v1.23.0

This version contains a minor fix that affects remote federated search users. If you are not a remote federated search user, it is not necessary to migrate from v1.22.x.

• v1.22.2 raised the remote federated search timeout for waiting nodes from 5 to 30s. This version makes it configurable by setting the environment variable MEILI_EXPERIMENTAL_REMOTE_SEARCH_TIMEOUT_SECONDS to a positive integer value. Please note that no CLI flag or configuration entry is available. By @dureuill in https://github.com/meilisearch/meilisearch/pull/5932