The Best LLMs in 2024

The best LLMs notable, intriguing, and widely used large language models are available today.

Large language models (LLMs) are the primary type of text-handling AIs and are becoming increasingly prevalent. ChatGPT is the most well-known tool utilizing an LLM, powered by a specially tuned version of OpenAI’s GPT models. 

However, many other chatbots and text generators, such as Google Gemini, and Anthropic’s Claude, also rely on LLMs.

Best LLMs in 2024

There are dozens of major LLMs and hundreds more significant for various reasons. Listing them all would be nearly impossible, and such a list would be outdated within days due to the rapid pace of LLM development. 

For example, updating this list just a few months after it was first published has already revealed new versions of multiple models and at least one new addition.

GPT

• Developer: OpenAI
• Access: API

OpenAI’s Generative Pre-trained Transformer (GPT) models ignited the latest AI hype cycle. 

Currently, four main models are available: GPT-3.5-turbo, GPT-4, GPT -4-turbo, and a new multimodal version called GPT-4o. 

These versions of GPT are general-purpose AI models with an API, and they are utilized by a diverse range of companies — including Microsoft, Duolingo, Stripe, and Dropbox — to power numerous tools. Despite this widespread use, ChatGPT remains the most popular demonstration of GPT’s capabilities.

Gemini

• Developer: Google
• Access: API

Google Gemini is a family of AI models from Google, consisting of three versions: Gemini Nano, Gemini Pro, and Gemini Ultra. These models are designed to operate on different devices, ranging from smartphones to dedicated servers. While they can generate text like an LLM, the Gemini models are also natively capable of handling images, audio, video, code, and other types of information.

Gemini Pro powers AI features throughout Google’s apps, such as Docs and Gmail, as well as Google’s chatbot, also named Gemini (formerly Bard). Gemini Pro 1.5 is available to developers through Google AI Studio or Vertex AI, with Gemini Nano and Ultra expected to be released later in 2024.

Google Gemma

• Developer: Google
• Parameters: 2 billion and 7 billion
• Access: Open

Google Gemma is a family of open AI models from Google, developed based on the same research and technology as the Gemini models. The Gemma models come in two sizes: 2 billion parameters and 7 billion parameters.

Llama 3

• Developer: Meta
• Parameters: 8 billion, 70 billion, and 400 billion (unreleased)
• Access: Open

Llama 3 is a family of open LLMs from Meta, the parent company of Facebook and Instagram. It powers many AI features across Meta’s apps and is one of the most popular and powerful open LLMs. The source code is available on GitHub, and it is free for research and commercial use, making it a common base for other LLMs.

Currently, Llama 3 offers versions with 8 billion and 70 billion parameters, with a 400 billion parameter version still in training. Meta’s previous model family, Llama 2, is also available in 7 billion, 13 billion, and 70 billion parameter versions.

Claude 3

• Developer: Anthropic
• Parameters: Unknown
• Access: API

Claude 3 is one of the most significant competitors to GPT. Its three models — Haiku, Sonnet, and Opus — are designed to be helpful, honest, harmless, and crucially, safe for enterprise customers to use. This has led to partnerships with companies like Slack, Notion, and Zoom.

Like other proprietary LLMs, Claude 3 is available only as an API, but it can be further trained on your data and fine-tuned to meet specific needs. Additionally, you can connect Claude to Zapier, allowing you to automate Claude’s functions across your other apps.

Falcon

• Developer: Technology Innovation Institute
• Parameters: 1.3 billion, 7.5 billion, 40 billion, and 180 billion
• Access: Open

Falcon is a family of open LLMs that have consistently excelled in various AI benchmarks. With models featuring up to 180 billion parameters, Falcon can outperform older models like GPT-3.5 in certain tasks.

Mixtral 8x7B and 8x22B

• Developer: Mistral
• Parameters: 45 billion and 141 billion
• Access: Open

Mistral’s Mixtral 8x7B and 8x22B models use a series of sub-systems to efficiently outperform larger models. Despite having significantly fewer parameters, enabling them to run faster or on less powerful hardware, they can surpass other models like Llama 2 and GPT-3.5 in some benchmarks. These models are released under an Apache 2.0 license.

Additionally, Mistral has released a more direct GPT competitor called Mistral Large, available through cloud computing platforms.

Grok

• Developer: xAI
• Parameters: Unknown
• Access: Chatbot and open

Grok, an AI model and chatbot trained on data from X (formerly Twitter), doesn’t warrant a place on this list based on its performance or popularity. However, it is notable because it was developed by xAI, the AI company founded by Elon Musk. Although it hasn’t made a significant impact in the AI scene, it receives substantial media coverage, making it worth mentioning.