Meili SAS

license

MIT(core)

enterprise-dir

BUSL-1.1

source

public

core features withheld

yes — advanced features including sharding and S3-streaming snapshots ship in the same repository under the Business Source License 1.1 or a commercial license, and BUSL permits non-production use only without an agreement

The LICENSE body reads `SPDX-License-Identifier: MIT AND BUSL-1.1`, splitting the repository in two: the core search engine is MIT and free for commercial use, while an Enterprise Edition portion covering sharding and S3-streaming snapshots is governed by BUSL-1.1. That functionality sits inside the published tree but outside the open terms, which is what gates the core.

• https://raw.githubusercontent.com/meilisearch/meilisearch/main/LICENSE recorded 2026-09-30

  LICENSE reads "`SPDX-License-Identifier: MIT AND BUSL-1.1`" and says "Part of this work fall under the Meilisearch Enterprise Edition (EE)".

• https://raw.githubusercontent.com/meilisearch/meilisearch/main/LICENSE-EE recorded 2026-09-30

  LICENSE-EE grants "You may use, modify, and distribute the Licensed Work for non-production purposes only, such as testing, development, or evaluation." and says "Production use of the Licensed Work requires a commercial license agreement with Meilisearch."

• https://raw.githubusercontent.com/meilisearch/meilisearch/main/README.md recorded 2026-09-30

  README license section lists Community Edition as "Fully open source under the [MIT license](./LICENSE)" and Enterprise Edition features "Sharding" and "S3-streaming snapshots" as "Not allowed in production without a commercial agreement with Meilisearch."

• https://www.meilisearch.com/docs/learn/self_hosted/enterprise_edition recorded 2026-09-30

  Docs page says "It is available under a BUSL license and cannot be freely used in production." and "The Enterprise Edition (EE) is a version of Meilisearch with advanced features."

The meilisearch package on PyPI is the Python client, not the engine, so its downloads measure client installs rather than server deployments. With no server deployment count published, GitHub stars are the adoption signal instead.

• https://api.github.com/repos/meilisearch/meilisearch recorded 2026-08-18

  Repo metadata - stargazers_count = 59,006 - for meilisearch/meilisearch.

Meilisearch is a complete search engine with typo-tolerant full-text, faceted, filtered, geosearch and hybrid semantic search, with embeddings produced by third-party models in the community edition and sharding in its enterprise portion. It sits one step below Vespa, which also hosts and evaluates models inside its serving path.

• https://raw.githubusercontent.com/meilisearch/meilisearch/main/README.md recorded 2026-09-30

  README feature list says "scale horizontally by distributing your data across multiple nodes" and "get relevant matches even when queries contain typos and misspellings" and "filter and sort documents based on geographic data".

• https://www.meilisearch.com/docs/learn/ai_powered_search/getting_started_with_ai_search recorded 2026-09-30

  Getting-started page says "Meilisearch relies on third-party embedding models to generate embeddings." and "Meilisearch will automatically batch your documents and send them to OpenAI for embedding generation."

• https://www.meilisearch.com/docs/learn/self_hosted/enterprise_edition recorded 2026-09-30

  Docs page says "It is available under a BUSL license and cannot be freely used in production." and "The Enterprise Edition (EE) is a version of Meilisearch with advanced features."