Anthropic is one of the leading AI companies in the world, and its models represent some of the biggest competition for OpenAI and Google. The company just announced its latest duo of AI models, and one of them may have just surpassed the competition. Anthropic just unveiled its Claude Opus 4 and Claude Sonnet 4 models.

Just like all other AI companies, Anthropic is working on AI agents that can work for you. This goes far beyond simply using a chatbot. The company wants to create a full-fledged AI assistant that can help you through your day. That’s one main focus point of Claude Opus 4. Also, this model will give the competition a run for its money and another area.

Anthropic unveiled Claude Opus 4 and Claude Sonnet 4

Starting off with the cheaper and more efficient model, Claude 4 Sonnet will be suited for more basic and general tasks. People will want to use this model if they want tasks and computations done quickly with a decent level of proficiency. It has a lower input and output cost, so companies using the API for this model won’t have to pay too much. 

Claude 4 sonnet scored pretty highly on some AI benchmarks, especially when it comes to coding. Anthropic says that the model delivers “superior coding and reasoning.” In fact, it beat out the competition on several AI benchmarks. You will see it and Claude 4 Opus highlighted on the chart below. Obviously, this is a chart provided by Anthropic itself, so there’s always the chance that the company is only highlighting the benchmarks that show its models in a good light.

Next, there is Claude 4 opus

This is a fitting name, as this model is Anthropic’s magnum opus. Claude Opus 4 is the most powerful model that the company has to date, and it brings some pretty significant reasoning capabilities. Just like Claude sonnet 4, Opus is geared toward coding. Anthropic states that Claude 4 Opus is the “best coding model in the world.” The fact that it surpasses Gemini 2.5 Pro is a momentous feat. Google put a specific emphasis on coding with its Gemini models.

Claude Opus 4 will be a force to be reckoned with when it comes to agentic tasks. The company announced that the model was able to run autonomously for several hours performing agentic tasks. Autonomy is something that AI companies are striving for, as that is one of the core features of AGI (artificial general intelligence). So, if Claude Opus 4 can run for several hours without any human intervention, then Anthropic could be the closest company to crafting a true agentic AI model.

If you want to try out these models, you are going to have to pony up in most cases. You can access Claude Opus 4 and Claude sonnet 4 through Amazon Bedrock, Google Cloud’s Vortex AI, Anthropic’s own API, and through the paid Claude plans. Also, for the time being, free users can use Claude Sonnet 4.