The Open-Source AI Stack
RSS

Glossary

source-available

A license category that lets users read and modify the code or weights but imposes restrictions (use limits, non-compete, MAU thresholds) that exclude it from the strict open-source definition.

Weights also: Governance aka source available

A license category that sits between fully open and fully proprietary. The source (or the weights, for models) are downloadable and editable. The license imposes restrictions that the OSIgovernanceThe nonprofit that maintains the canonical Open Source Definition for software since 1998, and the OSAID definition for AI as of 2024. Open full entry rejects as non-open-source: field-of-usegovernanceLicense clauses that limit which industries or applications a model may be deployed in, restricting use to non-competitive, non-commercial, or non-government purposes. Open full entry clauses, MAUgovernanceA user-count metric used in restrictive open-weights licenses (notably Llama's Community License) to trigger a requirement to negotiate a separate commercial license at scale. Open full entry thresholds, non-compete clauses, license-revocation hooks.

For models, this is how most Meta, Google, and Microsoft open releases work. LlamaweightsMeta's open-weight model family, the most widely deployed open release through 2024 to 2026, released under the source-available Community License with an MAU cap and acceptable-use clause. Open full entry ’s Community License with the 700M MAU cap is source- available, not open source. GemmaweightsGoogle's open-weight model family derived from Gemini research, with source-available licensing that includes an acceptable-use clause and license-revocation hook. Open full entry ’s terms with the acceptable-usegovernanceLicense or terms-of-service clauses that prohibit certain uses (weapons, surveillance, harassment, child sexual abuse material), common on open-weight licenses but rejected by the strict open-source definition. Open full entry revocation hook are source-available. The technique is to capture the PR benefit of “open” while retaining the legal levers of proprietary.

For an end user the practical effect is conditional permission. A hobbyist or small startup gets effectively-open access. A competitor or a large enterprise gets a contract negotiation. The mental model is not “free for everyone” but “free for the licensor’s chosen field of use.”

Sources

Mentioned in

Back to glossary