Discover the Game-Changing Features of Llama 3.1

Llama 3.1 raises AI standards with 128K context length and 405B model, challenging top closed-source models and enhancing open-source capabilities.

Llama 3.1 extends context length to 128K and comes in 8B, 70B, and 405B versions, raising the competition bar for large models.

For the AI community, the most significant aspect of Llama 3.1 405B is its new benchmark for open-source models. Meta claims it performs comparably to the best closed-source models in various tasks.

The table below shows the performance of the Llama 3 series on key benchmarks. The 405B model is close to GPT-4o in performance.