Anthropic, a Silicon Valley business supported by Google and Amazon.com announces a collection of artificial intelligence models under the moniker Claude 3. This is the most recent round in the nearly ceaseless competition to bring even more potent technologies to market.

The company claims that Claude 3 Opus, the most powerful member of the family, scores better on a variety of benchmark tests than competing models GPT-4 from OpenAI and Gemini 1.0 Ultra from Google.

Following numerous competitor announcements and the July release of Anthropic's Claude 2, the news demonstrates how businesses are competing to be at the top of charts ranking AI performance while business clients are still unsure about how to employ such technology.

Anthropic stated that it would charge $15 for Claude 3 Opus to process each of the one million tokens that are generated and that its smaller models would process the same data for at least five times less. In comparison, each million tokens that are entered into OpenAI's GPT-4 Turbo model costs $10.

Claude 3 is Anthropic's first "multimodal" AI package, according to the company. This indicates that Anthropic's AI can react to text inquiries and to pictures, such as assessing a picture or chart, just like some other competing models. Such generative AI, made popular by OpenAI's ChatGPT, produces new content.

Google stopped generating images for its Gemini chatbot last month because the service made mistakes on diversity, such as incorrectly portraying Nazi soldiers from the 1940s as being of different racial backgrounds instead of white.

CEO Amodei says, “Claude 3 will analyze photos but not make them as we've seen much less enterprise demand for them."

Anthropic said that the cloud computing services of Google and Amazon would offer their Claude 3 models. According to Anthropic, it would also directly sell access to them in 159 nations.