On Thursday (20th local time), OpenAI's main competitor, Antiopic, released the most powerful artificial intelligence model to date - the Claude 3.5 Sonnet.
It is reported that it can better understand complex instructions, improve visual reasoning abilities such as interpreting charts and graphics, and has a stronger sense of humor. It is worth mentioning that its app processing speed is twice that of the previous generation model Claude 3 Opus, and its cost is only one-fifth of the latter.
The company stated in a blog post, "It has shown significant progress in grasping subtle differences, humor, and complex explanations, and excels in writing high-quality content in a natural and relevant tone. It can also write, edit, and execute code."
Overall, in multiple performance tests covering areas such as reading, programming, mathematics, and vision, the Claude 3.5 Sonnet not only significantly improved its performance compared to its own Claude 3 Opus, but also slightly outperformed GPT-4o and Google's Gemini 1.5 Pro in multiple data points.
Like OpenAI's ChatGPT and Google Gemini, Claude has also been very popular in the past year. Anthropic was founded by a former OpenAI R&D executive and has gained support from Google, Salesforce, and Amazon. In the past year, Antiopic has completed five financing transactions, totaling approximately $7.3 billion.
Prior to this, Anthropic first launched the Claude 3 series model in March, while OpenAI launched the GPT-4o in May. The company stated that the Claude 3.5 Sonnet is faster than its previous leading model, the Claude 3 Opus, and is the first model in the newly launched Claude 3.5 series by Antiopic.
Affordable price
In terms of cost, Anthropic stated on its official website that Claude 3.5 Sonnet is currently available for free on the company's website Claude.ai and Claude iPhone application. Subscribers to Claude Pro and Team can also restrict access to the latest models at a higher rate. In addition, it is also provided through the Antitopic API, Amazon Bedlock, and Google Cloud's Vertex AI.
In addition, the new model is priced the same as the previous generation 3 Sonnet model, which costs $3 per million tokens input and $15 per million tokens generated, and has a context window of 200000 tokens, equivalent to approximately 150000 words.
In natural language processing (NLP), a token refers to the smallest semantic unit in a text. For example, a sentence can be divided into several words, each word being a token. For example, the sentence "I love you" can be divided into three tokens: "I", "love", and "you". Tokens can help us break down text into easier to process and analyze parts.
New features
At the same time as releasing the new model, Aerospace also introduced "Artifacts" on Claude.ai, a new feature that can expand the way users interact with Claude.
The company explained that when users request Claude to generate code snippets, text documents, or website designs, these artifacts will appear in a dedicated window next to their conversation. This creates a dynamic workspace where users can view, edit, and build Claude's creations in real-time, and seamlessly integrate AI generated content into their projects and workflows.
Anthropic also stated that it is expected that Artifacts will contribute to code development, legal contract drafting and analysis, and business report writing.
Nowadays, with the rise of start-ups such as Anthropic and OpenAI in the field of generative AI, they have become part of the AI arms race alongside tech giants such as Google, Amazon, Microsoft, and Meta. Experts predict that the market size will exceed $1 trillion in the next decade.