This site is fictional demo content. It is not real news or affiliated with any real organization. Do not treat it as fact or professional advice.

Full article

FULL TEXT

View this issue
BriefAI

Meta Open-Sources LLaMA 4: 50% Fewer Parameters, Same Performance, AI Open-Source Community Enthusiastic Again

Meta releases LLaMA 4 open-source large model, reducing parameter count by 50% while maintaining performance, significantly lowering inference costs, enabling SMEs to deploy private large models.

Meta today released LLaMA 4 open-source large model.

Performance

  • Parameter count: Reduced 50%
  • Performance: On par with LLaMA 3
  • Inference speed: 40% faster
  • VRAM requirement: Reduced 35%

Open Source License

  • Academic research: Completely free
  • Commercial use: Authorization required for revenue exceeding $5M

Market Impact

Open-source large models continue closing the performance gap with closed-source, and AI application development costs will further decline.