Amazon-backed Antropic has released Claude 4, touted as its most powerful group of AI models yet, made to tackle complex tasks.
The company announced on Thursday the release of the two models called Claude Opus 4 and Claude Sonnet 4 that are capable of analysing “thousands of data sources, executing long-running tasks,” as well as writing human-quality content.
Anthropic’s models score high on certain benchmarks
According to the AI startup, the new models score high on some popular benchmarks, showing that Opus 4 outperformed Google’s Gemini 2.5 Pro, OpenAI’s o3 reasoning, and GPT-4.1 models in coding tasks and using “tools” like web search.
Both models were tuned to perform well on programming tasks, making them suitable for writing and editing.
The company’s paying users and those who use free chatbot apps can both access the Sonnet 4, while the Opus 4 is a preserve of the paying customers only.
Opus 4 will be priced at $15/$75 per million tokens (input/output) and Sonnet 4 at $3/$15 per million tokens (input/output).
Tokens are the raw bits of data that AI models work with, with a million tokens equivalent to about 750,000 words – about 163,000 words longer than “War and Peace.”
Anthropic’s chief science officer, Jared Kaplan, revealed that the company stopped investing in chatbots at the end of 2024 but has been focused on improving Claude’s capabilities to handle complicated tasks, including research and coding.
Kaplan admitted that “the complex the task is, the more risk there is that the model is going to kind of go off the rails…and we’re really focused on addressing that so that people can really delegate a lot of work at once to our models.”
“I think these models are much, much stronger as agents and as coders. It was definitely a struggle internally just because some of the new infrastructure we were using to train these models… made it very down-to-the-wire for the teams in terms of getting everything up and running.”
Kaplan.
Anthropic boasts the Opus 4 model, referring to it as the “best coding model in the world” and could autonomously work for nearly a full corporate workday – hours.
Anthropic added “thinking summaries” for both models
Founded by former OpenAI research executives, Anthropic debuted its first chatbot – Claude, in March 2023, posing pressure on OpenAI. Since that launch, the company has been part of the fierce battle in the cut-throat AI industry, especially a battle of startups.
This also comes as businesses across sectors are rushing to add AI-powered chatbots and agents to stay relevant in the game and not be left behind by competitors.
Now, for Anthropic, the company added a new feature for both Claude 4 models, described as “thinking summaries.” This condenses the chatbot’s reasoning process into easily understandable insights.
According to The Verge, the “extended thinking” feature is also launching in beta, which allows users to switch models between modes for reasoning or using tools to improve the accuracy of responses.
The models also have the capacity to search the web to complete tasks on a user’s behalf and alternate between reasoning and tool using. When given access to local files, the models can extract and save “key facts to maintain continuity and build tacit knowledge over time.
“I do a lot of writing with Claude, and I think prior to Opus 4 and Sonnet 4, I was mostly using the models as a thinking partner, but still doing most of the writing myself.”
Anthropic chief product officer Mike Krieger.
“And they’ve crossed this threshold where now most of my writing is actually … Opus mostly, and it now is unrecognizable from my writing,” added Krieger.
The models are available on the Anthropic API, Amazon Bedrock, and Google Cloud’s Vertex AI platform.
