Why Llama 3.1 405B Is So Much Better Than GPT-4o And Claude 3.5 Sonnet— Here The Result

Gao Dalie (高達烈)
Towards AI
Published in
11 min readJul 26, 2024

AI news in the past 7 days has been insane, with so much happening in the world of AI

In this Story, we’re diving into some of the latest AI developments from major players like Llama 3.1 405B, GPT-4o, and Claude 3.5 Sonnet

Llama 3.1 405B is the first open-source model that performs on par with leading proprietary AI models in general knowledge, steerability, math, tool use, and multilingual translation, among other capabilities.

Meta announced the launch of Llama 3.1, which is the largest open-source AI model to date and has surpassed OpenAI’s GPT-4o and Anthropic’s Claude 3.5 Sonnet in multiple benchmark tests!

In this step-by-step guide, we will cover what Llama 3.1 405B is, how to use Llama 3.1 405B locally, and why Llama 3.1 405B is so much better than GPT-4o and Claude 3.5 Sonnet.

I highly recommend you watch this video to the end is a game changer in your chatbot that will realize the power of Llama 3.1 405B!

Published in Towards AI

The leading AI community and content platform focused on making AI accessible to all. Check out our new course platform: https://academy.towardsai.net/courses/beginner-to-advanced-llm-dev

Written by Gao Dalie (高達烈)

Learn AI Agent, LLMs, RAG & Generative AI See everything I have to offer at the link below: https://linktr.ee/GaoDalie_AI

Responses (10)

What are your thoughts?

The title is a bit misleading though - since the title said WHY, I expect a deep-down analysis of WHAT meta did right that results in such a great improvement that, instead of HOW 3.1405B is comparable to the closed-source models.

--

trained it on more than 16,000 H100 GPUs

do you know what minimum hardware you need for inference with 3.1 405B? e.g. just using the model not retuning it.

--

Thanks for the insightful article. As a European, I'm interested in using Llama 3.1, but it seems Meta is completely dismissive of our ethical and data privacy standards. They're not offering Llama in Europe at all because they ignore the ethical guidelines set by the EU AI Act. Is this not a concern in the US?

--