Meta AI Introduces Llama 3.1 405B, the Most Powerful Open-Source AI Model in Human History
Meta AI’s new artificial intelligence model, Llama 3.1 405B, has been introduced. This model is the most powerful one available to the public so far and is free to use.
The battle of AI applications continues. In the competition involving companies like OpenAI and Microsoft, Meta has suddenly taken the lead with its new model. The company’s newly introduced model, Llama 3.1 405B, surpasses all its competitors in many areas.
The new Llama comes in three different sizes. The 8B model can run on any device with 8GB of RAM. The 70B model requires at least 64GB of RAM, while the 405B model is the most powerful and will be accessible through service providers.
GPT-4o and Claude 3.5 Lead Ahead of Sonnet
The 405B model competes head-to-head with the market’s most powerful closed-source AI models, GPT-4o and Claude 3.5 Sonnet, in tasks such as general information, navigability, mathematics, tool usage, and multilingual translation. Smaller models also show performance on par with GPT-4o and Claude 3.5 Sonnet.
Llama 3.1 405B, trained on NVIDIA’s H100 GPU for 39 million hours, offers Llama Stack API support for easy integration. With this model, open-source AI language models have caught up with their closed-source, paid counterparts. It remains to be seen how companies like OpenAI will respond to this model.
With the introduction of this model, administrations, institutions, companies, and individual users will have free access to the most advanced AI model.
