# Open-Weight LLM License Terms Reference > Source: https://netray.co/tools/open-weight-llm-license-terms-reference > Publisher: Netray (netray.co), AI and integration engineering for ERP and manufacturing systems. > Section: Free Tools ## Short answer Look up which license each popular open-weight model family ships under and what to check before running it on-prem against ERP data. ## Category on-prem-ai ## Audience - CIO - IT manager - Legal counsel - ERP architect ## Tool **Kind:** worksheet **Mode:** reference **Columns:** #### Model family **Key:** family **Type:** text #### Publisher **Key:** vendor **Type:** text #### License (typical) **Key:** license **Type:** text #### Commercial use **Key:** commercial **Type:** text #### Conditions to check **Key:** conditions **Type:** text #### On-prem ERP note **Key:** erp **Type:** text **Rows:** **Family:** Llama 3.1 / 3.3 **Vendor:** Meta **License:** Llama Community License **Commercial:** Yes, with conditions **Conditions:** Attribution (Built with Llama), acceptable use policy, separate license needed above 700M monthly active users, derivative model naming rule **Erp:** Fine for internal ERP assistants; keep the AUP and attribution text in your documentation **Family:** Llama 4 **Vendor:** Meta **License:** Llama 4 Community License **Commercial:** Yes, with conditions **Conditions:** Similar structure to earlier Llama licenses; read the current text for any regional restrictions on multimodal use **Erp:** Have counsel read the exact license for your country before deploying **Family:** Qwen3 (most sizes) **Vendor:** Alibaba Qwen **License:** Apache 2.0 **Commercial:** Yes **Conditions:** Keep license and notices; verify the model card for each size **Erp:** Permissive; common choice for air-gapped deployments **Family:** Qwen2.5 (mixed) **Vendor:** Alibaba Qwen **License:** Apache 2.0 for most sizes; some sizes use the Qwen license or a research license **Commercial:** Depends on size **Conditions:** Check the LICENSE file of the exact checkpoint, since sizes differ **Erp:** Record the exact checkpoint and its license in your model inventory **Family:** Mistral 7B **Vendor:** Mistral AI **License:** Apache 2.0 **Commercial:** Yes **Conditions:** Keep license and notices **Erp:** Small and cheap to host; limited on long ERP documents **Family:** Mixtral 8x7B **Vendor:** Mistral AI **License:** Apache 2.0 **Commercial:** Yes **Conditions:** Keep license and notices **Erp:** Mixture of experts, needs memory for all experts **Family:** Mistral Small (recent open releases) **Vendor:** Mistral AI **License:** Apache 2.0 **Commercial:** Yes **Conditions:** Verify per release; some Mistral models use other licenses **Erp:** Good size for a single-GPU private assistant **Family:** Mistral Large 2 **Vendor:** Mistral AI **License:** Mistral Research License **Commercial:** No, research use only without a commercial agreement **Conditions:** Commercial deployment needs a separate agreement with the publisher **Erp:** Do not deploy in production without that agreement **Family:** Codestral 22B **Vendor:** Mistral AI **License:** Mistral AI Non-Production License **Commercial:** No for production **Conditions:** Non-production and research use only unless licensed separately **Erp:** Check before using for ABAP or SQL code generation inside a company **Family:** Gemma 2 / Gemma 3 **Vendor:** Google **License:** Gemma Terms of Use **Commercial:** Yes, with conditions **Conditions:** Prohibited use policy applies and must be passed to downstream users; not an OSI open source license **Erp:** Fine for internal use; review redistribution terms if you ship it to customers **Family:** gpt-oss-20b / gpt-oss-120b **Vendor:** OpenAI **License:** Apache 2.0 with a usage policy **Commercial:** Yes **Conditions:** Keep license; the accompanying usage policy applies **Erp:** Open-weight reasoning models that run on-prem; check VRAM needs first **Family:** DeepSeek-R1 **Vendor:** DeepSeek **License:** MIT **Commercial:** Yes **Conditions:** Keep license notice. Distilled variants inherit the license of the base model (Qwen or Llama) **Erp:** Check the base model license for any distilled checkpoint you use **Family:** DeepSeek-V3 **Vendor:** DeepSeek **License:** Code under MIT; weights under the DeepSeek Model License **Commercial:** Yes, with use restrictions **Conditions:** Model license lists prohibited uses **Erp:** Very large; needs a multi-GPU node **Family:** Phi-3 / Phi-4 **Vendor:** Microsoft **License:** MIT **Commercial:** Yes **Conditions:** Keep license notice **Erp:** Small models suited to edge and plant-floor devices **Family:** Granite **Vendor:** IBM **License:** Apache 2.0 **Commercial:** Yes **Conditions:** Keep license and notices **Erp:** Enterprise-oriented small and mid-size models **Family:** OLMo **Vendor:** Allen Institute for AI **License:** Apache 2.0 **Commercial:** Yes **Conditions:** Keep license and notices **Erp:** Fully open training data and code, useful where auditability matters **Family:** Command R / R+ (open weights) **Vendor:** Cohere **License:** CC BY-NC 4.0 with acceptable use addendum **Commercial:** No for non-commercial license **Conditions:** Commercial use requires a commercial agreement **Erp:** Not usable for production without that agreement **Family:** Nemotron **Vendor:** NVIDIA **License:** NVIDIA Open Model License **Commercial:** Yes, with conditions **Conditions:** Read the current license for guardrail and redistribution terms **Erp:** Pairs naturally with NVIDIA inference stacks **Notes:** Licenses change between releases and between sizes of the same family. Always read the LICENSE file shipped with the exact checkpoint you download. This table is a starting point for your review, not legal advice. ## Content **Intro:** Before an open-weight model touches your ERP data, someone has to confirm you are allowed to run it for commercial work. This reference lists the license typically attached to each popular family and the conditions people most often miss. **Sections:** #### How to use this table **Body:** Search by model family or by license name. Read the Commercial use column first to rule models out, then the conditions column for attribution, acceptable use policies and user-count thresholds. Add the exact checkpoint name and license to your internal model inventory so that an auditor or a customer questionnaire can be answered in minutes. Export the filtered rows to Excel or PDF to attach to a procurement or architecture decision record. #### Open weights are not the same as open source **Body:** A model with downloadable weights can still carry restrictions. Apache 2.0 and MIT are permissive. The Llama and Gemma licenses are custom terms with acceptable use policies. Research or non-production licenses forbid use in a live business process. The distinction matters because an internal ERP assistant that answers real purchasing or production questions is a commercial use even if no customer ever sees it. **Bullets:** - Permissive: Apache 2.0, MIT - Custom community licenses: attribution and policy flow-down - Non-commercial or research only: needs a separate agreement - Distilled models inherit the base model license #### Common mistakes **Body:** Teams often judge a family by one famous release and assume every size shares its license. Others forget that a fine-tuned derivative keeps the obligations of its parent. A third mistake is ignoring the acceptable use policy, which can prohibit use cases that touch regulated areas. Finally, license text can be updated in later releases, so a version you approved last year may not match the one you download today. **Cta:** Netray deploys open-weight models on customer GPUs and can help you document the license, hardware and data flow for each one. We do not give legal advice, so counsel should confirm the final reading. **Faq:** #### Can I use Llama commercially for an internal ERP assistant? **Answer:** Generally yes under the Llama Community License, subject to its conditions such as the acceptable use policy and attribution. The 700 million monthly active user threshold rarely matters for a company-internal tool. Read the exact text for the release you use. #### Does Apache 2.0 mean there are no restrictions at all? **Answer:** No. Apache 2.0 allows commercial use, modification and redistribution, but you must keep the license and notice files and state significant changes. It also includes a patent grant. Some models add a separate usage policy on top of it. #### Are distilled DeepSeek models covered by the MIT license? **Answer:** The DeepSeek-R1 release is MIT, but distilled checkpoints are built on Qwen or Llama bases and also inherit those base licenses. Check both before approving a specific checkpoint for production. #### How often should we recheck licenses? **Answer:** At each model upgrade and at least annually. Publishers sometimes change terms between versions. Store the license file and download date alongside the model weights in your artifact repository. ## Search terms open weight llm license comparison, llama community license commercial use, apache 2.0 llm models list, can i use qwen commercially, gemma terms of use commercial, gpt-oss license, deepseek r1 mit license, open source llm license table --- Talk to an engineer: https://netray.co/contact?intent=erp_ai | Machine catalog: https://netray.co/agent/catalog.json