Model weights license
The model weights license defines what you can / can't do with downloaded model weights — Apache 2.0 (permissive), MIT (permissive), Llama Community License (restricted > 700M MAU), Gemma terms, FLUX-1-Dev non-commercial — production-critical to read before deploying.
Open-weight does not mean license-free. Common 2026 license patterns: Apache 2.0 / MIT (truly permissive — Mistral, Qwen, Phi) with no commercial restrictions; Llama Community License (free up to 700M MAU, then requires Meta license); Gemma terms (Google-specific use restrictions); FLUX-1-Dev (non-commercial), FLUX-1-Schnell (Apache 2.0); Stable Diffusion 3 Medium (community license with revenue limits); custom 'Responsible AI' licenses with use-case restrictions. Production gotcha: legal must review before shipping; license terms can change between versions; some licenses prohibit competitive use (training models on outputs). Read the LICENSE file before integrating.
When to use model weights license
- Before shipping any open-weight model in production.
Common mistakes
- Skipping the LICENSE file — assuming permissive when it isn't.
- Not tracking license changes between versions — a v2 may have stricter terms than v1.
FAQ
What is model weights license?
The model weights license defines what you can / can't do with downloaded model weights — Apache 2.0 (permissive), MIT (permissive), Llama Community License (restricted > 700M MAU), Gemma terms, FLUX-1-Dev non-commercial — production-critical to read before deploying.
When should I use model weights license?
Before shipping any open-weight model in production.
What are the most common mistakes with model weights license?
Skipping the LICENSE file — assuming permissive when it isn't. Not tracking license changes between versions — a v2 may have stricter terms than v1.
Related terms
- Open-weight model — An open-weight model has publicly released weights downloadable + runnable by anyone — Llama, Mistral, Qwen, DeepSeek, Flux Schnell / Dev, Stable Diffusion are 2026 open-weight families. Differs from open source (which would include training code + data).
- Self-hosted LLM — A self-hosted LLM runs entirely on infrastructure you control — your GPUs, your servers, your data residency — versus calling a cloud API.
Last updated: 2026-06-01. Raw markdown: https://promtable.com/glossary/model-weights-license.md.