Meta Launches Llama 4 Model While Tightening Open‑Source Licensing Strategy for AI Developers

Meta announced the release of Llama 4, its latest large language model, positioning it as an open AI model designed for next‑generation applications, according to the company’s technical guide.

Zuckerberg said the company will limit the model’s public code distribution, marking a retreat from the fully open‑source stance that characterized earlier Llama versions, a move he described as a “sensible pivot.”

Llama 4 appears in a comparative review of fifteen tested open‑source AI models for 2026, where it ranked near the top for language understanding and inference speed, yet the same review notes the new licensing terms limit unrestricted commercial deployment.

Industry observers predict the tighter licensing could shift developer adoption toward alternative open models, while Meta hopes the change will protect intellectual property and improve monetization of AI services and expand its cloud AI offerings.