Open-source LLMs are large language models released publicly for anyone to use, modify, or deploy. They offer flexibility and transparency that proprietary models do not.
Organizations that need to run models on their own infrastructure, customize model behavior deeply, or avoid per-token API costs often choose open-source LLMs. The trade-off is that deployment and maintenance require more internal technical capability. Open-source does not mean zero cost — it shifts cost from licensing to infrastructure and engineering.