The Claude AI model of Anthropic can now manage longer prompts

Anthropic increases the amount of information that business customers can send to Claude in a single prompt, part of an effort to attract more developers to the popular Coding models of the company.

For Anthropic API customers, the company Claude SONNET 4 The AI model now has a context window of 1 million tokens – which means that AI can process requests up to 750,000 words, more than the entire “Lord of the Rings” trilogy, or 75,000 lines of code. It is about five times the previous limit of Claude (200,000 tokens), and more than double the context window of 400,000 tokens offered by the OPENAI GPT-5.

Long Context will also be available for Claude Sonnet 4 via Anthropic Cloud partners, including on Amazon Bedrock and Google Cloud’s Green AI.

Anthropic has built one of largest company Among the developers of AI models, largely by selling Claude to the AI coding platforms such as Microsoft’s Github co-pilot, Windsurf and the Anysphere cursor. While Claude has become the model of choice among the developers, the GPT-5 can threaten the domination of anthropic with its competitive prices and its strong coding performance. Michael Truell, CEO of Anysphere, even helped Openai Launch of GPT-5Who is now the default AI model for new users in the cursor.

Anthropic’s product manager for the Claude platform, Brad Abrams, told Techcrunch in an interview that he expects to expect him to expect AI coding platforms that he has a “many advantages” of this update. When asked if the GPT-5 had put a breach in the use of the API of Claude, Abrams minimized the concern, saying that it is “really satisfied with the API company and the way it develops”.

While Openai generates most of its income from Cons consumption subscriptions to Chatgpt, anthropic activities focus on the sale of AI models to companies via an API. This has made coding platforms a key customer for anthropogenic and could be the reason why the company launches new advantages to attract users against GPT-5.

Last week, Anthropic unveiled an updated version of its biggest AI model, Close work 4.1 which has pushed the coding capacities of the AI of the company a little further.

In general, AI models tend to work better on all tasks when they have more context, but especially for software engineering problems. For example, if you ask an AI model to run a new feature for your application, it is likely to do a better job if it can see the entire project, rather than a small section.

Abrams has also told Techcrunch that the large context window of Claude helps to better work on long agent coding tasks, in which the AI model works independently on a problem for minutes or hours. With a large context window, Claude can remember all of her previous steps in long-horizon tasks.

But some companies have taken large context windows to the extreme, saying that their AI models can process massive prompts. Google offers a context window of 2 million tokens for Gemini 2.5 Pro, and Meta offers a context window of 10 million tokens for Llama 4 Scout.

Some studies suggest that there is a limit to the effectiveness of large context windows; AI models are not great To treat these massive invites. Abrams said that Anthropic’s research team has focused on the increase not only of the Claude context window, but also of the “effective context window”, which suggests that its AI can understand most of the information given to it. However, he refused to reveal the exact anthropic techniques.

When the prompts to Claude Sonnet 4 are greater than 200,000 tokens, Anthropic will charge for API users, at $ 6 per million entry tokens and $ 22.50 per million exit tokens (compared to $ 3 per million entry tokens and $ 15 per million production tokens).

We always seek to evolve, and by providing an overview of your point of view and your comments on Techcrunch and our coverage and our events, you can help us! Fill this survey to let us know how we do it AAnd have the chance to win a prize in return!