Anthropic, a startup supported by Google and Amazon, recently introduced a set of powerful artificial intelligence models known as Claude 3. This announcement adds to the ongoing competition in Silicon Valley to develop increasingly sophisticated technology.
According to Anthropic, their most capable model, Claude 3 Opus, surpasses competitors like GPT-4 from OpenAI and Gemini 1.0 Ultra from Google in various benchmark exams. CEO Dario Amodei described it as the “Rolls-Royce of models” for the current period.
While advances from Anthropic’s competitors are expected, Reuters could not independently verify these claims. This news follows several other announcements in the tech industry and comes after Anthropic’s Claude 2 release in July. Companies are striving to lead AI performance charts while businesses are still figuring out how to effectively utilize such technology.
Anthropic’s President, Daniela Amodei, stated that customers would choose Claude 3 Opus, even with a higher price tag, for tasks requiring high cognitive complexity, such as accurate financial analysis.
The pricing for Claude 3 Opus is set at $15 for every 1 million pieces of data (tokens), which is at least five times less than its smaller models for the same task. In comparison, OpenAI charges $10 for each million tokens entered into its GPT-4 Turbo model.
Additionally, Anthropic highlighted that Claude 3 is its first “multimodal” AI suite. Similar to other models, Anthropic’s AI can respond to both text queries and images, such as analyzing photos or charts. CEO Amodei clarified that Claude 3 would examine images but not generate them due to less enterprise demand, addressing concerns raised by Google’s Gemini chatbot regarding inaccurate image generation.
Anthropic plans to make Claude 3 models available through Amazon and Google’s cloud platforms. They will also offer direct access in 159 countries.
Source: Reuters
