Anthropic launches Claude Sonnet 5.5 Anthropic has announced the release of Claude Sonnet 5.5, the newest member of its Claude 5.5 family, designed to deliver faster speed, less expensive operation and enhanced efficiency on common AI workloads. This new offering comes as businesses demand AI solutions capable of producing code, documents, research and workflow automation without the price of higher-end flagship models.
In addition, Anthropic said its Claude Sonnet 5.5 is also intended as a more cost-effective and faster partner to Claude Opus 5.5 for specific well-understood tasks that developers and businesses use frequently. Anthropic said the new model is more than 30% faster than the previous generation Claude Sonnet 5, and potentially up to 30% cheaper for typical workloads.
The main concern in Sonnet 5.5 is how quick and efficient
6 The performance of Claude Sonnet 5.5 “This is a highly efficient model,” Anthropic told Tech Crunch. “We’ve found that it can perform similar tasks with a smaller number of tokens compared to our previous model, which helps lower the total cost of AI-automated workflows.”
Claude Sonnet 5.5 is $2 per million input tokens, $10 per million output tokens (the same as Sonnet 5). Anthropic estimates the new model can cost up to 30% less to run for common use cases because it needs to process fewer tokens to do the same work. Prompt caching is estimated to save up to 90% of costs, batching up to 50%.
This method could be especially applicable to companies deploying AI at scale, where a tiny optimization can save a company thousands or millions of requests, and a tiny savings in computing time would add up.
Better AI coding capabilities
For Claude Sonnet 3.5, coding continues to be one of the most critical applications. The new model was created, according to Anthropic, to be well suited for things like building features, debugging code, performing code reviews, and working in existing codebases.
In the company’s tests, Sonnet 5.5 scored 70.6% on Terminal-Bench 4.0, versus 10.3% for Sonnet 5. In its announcement, Anthropic also noted gains in long-horizon tasks, in which an AI system has to retain focus over several steps in order to produce a final output.
To developers, the value is not in creating code. We envision the model as a way to gain insight into a codebase, roadmap modifications, perform incremental edits, and complete tasks with less reliance on calling the tool. This may lead to a reduction in developers’ time spent reviewing or fixing AI work.
Designed for everyday business tasks
Nlp Anthropic’s goal is to get Claude Sonnet 5.5 “outside of programming”. The company states that the model is well-suited to daily work tasks such as putting together documents, slides and spreadsheets.
Design and visual understanding upgrades The model can generate well-organized charts and user flow diagrams as well as adhere to presentation templates and deliver visually attractive output with minimal edits, in the words of the developers from Anthropic.
These features could expand how a company could use AI for other functions, rather than restrict it to engineering teams. Marketers, financial professionals, operators, customer support teams, and researchers could leverage the model to handle monotonous tasks, generate ideas, and work through various processes.
Claude Sonnet 5.5 availability
You can try Claude 5.5 via Claude.ai via web, iOS and Android. Developers can access Claude 5.5 via Anthropic’s Claude Platform. Currently, Anthropic indicates you can access it from major cloud infrastructure providers such as AWS, Google Cloud, and Microsoft Foundry.
The model is also available already via GitHub Copilot where GitHub says it has been focused on more well-scoped everyday development tasks, including building features and fixing bugs.
Anthropic expands its Claude 5.5 lineup
Sonnet 5.5 -which launched after Claude Opus 5.5- expands Anthropic’s range of models for different sizes of workloads. Opus 5.5 tends to be used for more demanding tasks that need more reasoning, while Sonnet 5.5 is more for speed, efficiency and day-to-day professional work.
Anthropic has also announced that it will soon be adding a new to its suite of Claude models named “Claude Haiku 5.5” into its Claude 5.5 suite of models for high-volume and low-cost use.
The launch signals a bigger trend across the AI ecosystem, where model creation is now not only evaluating benchmarks but the dollars, hours and tokens needed to do real-world work.
Anthropic hopes to marry a more powerful version of its AI with faster speed and a lower true cost of running it with Claude Sonnet 5.5. For large-scale organizations and developers deploying AI, that efficiency might be a more compelling concern when shopping around.
