What Is Claude 3? Anthropic’s New AI Models Explained

Lastly, there is Claude 3 Haiku, which is the third currently unreleased model. It is at the bottom of the family tree in terms of sophistication and most capabilities, but is designed more to be the fastest version and provide almost instant responses.

Haiku seems to be aimed almost exclusively at businesses and not everyday users. Its speed means it’s the model that will help power things like customer-facing AI chatbots for websites, which help you learn more about what you’re trying to buy, book appointments and resolve other queries or complaints.

It is also likely to be preferred by companies that need to digest large amounts of data and reports urgently.

Claude 3 Pricing: How Much Does Claude 3 Opus Cost?

As we have said, some models of Claude 3 require you to pay a subscription to Claude Pro. Specifically, that is the case of Claude 3 Opus, the most advanced version.

This effectively means that The price of Claude 3 Opus is set at $20 per month for consumers. Companies that want to use it will pay using the common token system for custom implementations of AI chatbot technology, which for Opus is $15 per million input tokens and $75 per million output tokens.

Claude 3 Sonnet is free to use in the open version of Claude AI that is available to users around the world, the “worker” model in Anthropic’s own words, and companies can make it their own for $3 per million entry tokens and $15 per million exit tokens.

At this time, the Claude 3 price for Haiku is only available in token form: $0.25 per million tokens in and $1.25 per million tokens out.

Claude 3 vs GPT-4: first look

We are still testing the capabilities of Claude 3 compared to key competitors such as ChatGPT’s GPT-4 model. As such, we are not yet in a position where we can share our thoughts.

That said, Anthropic has published a series of Claude 3 performance tests, which it says show how the Claude 3 family stacks up against the likes of ChatGPT and Google’s Gemini Ultra LLM.

Of course, the key caveat here is that these are Anthropic’s own tests and will rely, to a greater or lesser extent, on guidance heavily designed to take full advantage of Claude 3’s capabilities.

That’s not to say that Claude 3 isn’t enormously capable. It seems that it is. However, we can’t definitively say whether it’s better or worse than ChatGPT, Gemini, or any other chatbot until we’ve fully tested it ourselves.