Claude 3.5 Sonnet
Enhanced reasoning, state-of-the-art coding skills, computer use, and 200K context window
Announcements
- New
Claude 3.5 Haiku and a new Claude 3.5 Sonnet
Oct 22, 2024
Our updated version of Claude 3.5 Sonnet is now state-of-the-art for real-world software engineering tasks, agentic capabilities, and computer use in public beta.
Read more
- NEW
Developing a computer use model
Oct 22, 2024
A discussion of how our researchers developed Claude's new computer use skills, as well as some of the relevant safety considerations.
Read more
Availability and pricing
For developers interested in building custom AI solutions with Claude 3.5 Sonnet, it is available on the Anthropic API, Amazon Bedrock, and Google Cloud's Vertex AI.
For business users and consumers who want to collaborate with Claude 3.5 Sonnet using a simple chat experience, Claude 3.5 Sonnet is available on Claude.ai for all users across the web, iOS, and Android.
Pricing for Claude 3.5 Sonnet starts at $3 per million input tokens and $15 per million output tokens, with up to 90% cost savings with prompt caching and 50% cost savings with the Message Batches API. To learn more, check out our pricing page.
Use cases
Claude 3.5 Sonnet can understand nuanced instructions and context, recognize and correct its own mistakes, and can create sophisticated analysis and insights from complex data. Combined with state-of-the-art coding, vision, and writing skills, you can use Claude 3.5 Sonnet for a variety of use cases. Popular use cases include:
Computer use
By integrating Claude via API, developers can direct Claude to use computers the way people do—by looking at a screen, moving a cursor, clicking buttons, and typing text. Claude 3.5 Sonnet is the first frontier AI model to reliably use computers in this way—albeit experimentally in public beta—and we expect the capability to improve over time.
Code generation
Claude 3.5 Sonnet can help across the entire software development lifecycle—from initial design to bug fixes, maintenance to optimizations. Build Claude into your own products or use it as a coding assistant via Claude.ai.
Advanced chatbots
With enhanced reasoning and a warm, human-like tone, Claude 3.5 Sonnet is ideal for chatbots that need to connect data and take action across a variety of systems and tools.
Knowledge Q&A
Claude 3.5 Sonnet offers a large context window and low rates of hallucination, making it ideal for answering questions around large knowledge bases, documents, and codebases.
Visual data extraction
Claude 3.5 Sonnet is able to extract information from visuals like charts, graphs, and complex diagrams with ease—making it an ideal AI model for data analytics and data science tasks.
Robotic process automation
Automate repetitive tasks or processes with Claude 3.5 Sonnet. It offers industry-leading instruction following and is capable of handling complex processes and operations.
Benchmarks
Claude 3.5 Sonnet offers state-of-the-art performance across a variety of coding, vision, and reasoning tasks.
Trust & Safety
Anthropic's commitment to safety shapes every step of our AI development. During Claude 3.5 Sonnet's development, we conducted extensive safety evaluations spanning multiple languages and policy domains. We also enhanced Claude's ability to navigate sensitive content with appropriate care. Our internal testing confirms that Claude 3.5 Sonnet delivers substantial advances in capabilities while upholding our rigorous safety standards.
Computer use with Claude 3.5 Sonnet in Amazon Bedrock is helping us process diverse menu formats. We’re excited by its potential to consistently keep our menu database up-to-date, ensuring customers can experience the latest offerings. We’re creating more intuitive experiences and transforming how we handle voice orders with AI.
The new Claude 3.5 Sonnet shows significant improvements in handling complex instructions, creativity, and formatting. We’re impressed by its clearer text rephrasing and improved creative performance. These align with our goal of enhancing user productivity in Notion.
Claude 3.5 Sonnet sets a new standard for AI-assisted legal work. We’ve observed a 25% improvement on difficult contract labeling tasks and a 35% improvement in human acceptance of AI suggestions for editing. Its enhanced reasoning works across more contract types and complex legal scenarios.
We’ve been incredibly impressed with the new Claude 3.5 Sonnet’s ability to swiftly handle complex transcript analysis—condensing hours of work into seconds. It catches subtle contradictions across multiple depositions and drills down into issues that matter. The value-add to attorneys and their staff is immense.
We’ve reached a turning point where LLM capability is no longer the bottleneck for AI-assisted data science. Our focus is now on infrastructure for AI to autonomously train and evaluate models. The new Claude 3.5 Sonnet’s ability to visualize data and create charts—surpassing the previous version in quality and accuracy—is particularly impressive.
The new Claude 3.5 Sonnet is bringing us closer to our mission of creating a world where everyone can enjoy manga. By improving translation quality and speed, it enables more authentic, natural-sounding localizations across markets. This is transforming how quickly and effectively we can bring manga to global audiences.
The new Claude 3.5 Sonnet has significantly enhanced our AI compliance reviews for fintechs. It matches traditional OCR accuracy without initial processing and excels at visually comparing documents against templates, providing detailed authenticity reports. It produces human-like descriptions of discrepancies, improving our verification process and maintaining rigorous standards.
The new Claude's computer use goes above and beyond our expectations, and we’re ready to build a key capability of our Replit Agent based on it. By navigating new UIs with a thoughtful testing plan, it opens the door to creating a powerful autonomous verifier that can evaluate apps while they’re being built. This is the breakthrough we’ve been waiting for.
The new Claude 3.5 Sonnet’s ability to understand and interact with computers and browsers is opening exciting new frontiers for AI-assisted design at Canva. We’re eagerly exploring how this breakthrough might accelerate design processes and create time-savings that could be game-changing for users. This technology has great potential to empower the world to design.
With the new Claude 3.5 Sonnet, we’re entering a new era of AI-powered DevSecOps. Across our use cases, we’re seeing 5-10% improvements over the previous version. The new Claude 3.5 Sonnet’s careful reasoning comes without the usual latency hit. This makes it ideal for multi-step, agentic software development. We’re excited to use this industry-leading model in our next major AI-powered feature, Duo Workflow.
Anthropic is onto something special. The new Claude 3.5 Sonnet with computer use blew us away. It’s the best we’ve seen for automating web-based workflows, outperforming everything else we’ve tried. Its ability to navigate complex scenarios is remarkable. The rapid progress is incredibly exciting—we’re eager to see it keep getting better.
The new Claude 3.5 Sonnet with computer use is a leap forward in autonomous AI. It outperformed the previous version in our evaluations, showing marked improvements in coding, planning, and problem-solving. Its visual capabilities enable human-like UI interaction. Together, they’re opening doors to a new generation of AI products capable of real-world decision-making and action.
The upgraded Claude 3.5 Sonnet shows great potential for Asana AI - particularly as we continue expanding AI Studio to take on cross-tool actions on your behalf. Its reliable understanding of specific UI tasks and accurate interaction with complex interfaces is impressive
Frequently asked questions
We offer a family of Claude models across the spectrum of speed, price, and performance. Claude 3.5 Sonnet is our most intelligent model to date. We recommend Claude 3.5 Sonnet for critical use cases where you want frontier intelligence, like customer-facing AI experiences.
Pricing depends on how you want to use Claude 3.5 Sonnet. To learn more, check out our pricing page.
Instead of restricting Claude to use APIs, we're teaching it general computer skills—allowing it to use a wide range of standard tools and software programs. Developers can use this beta capability to automate repetitive tasks, perform software testing and Q/A, and perform open-ended tasks like research.
To make this possible, we've built an API that allows Claude to perceive and interact with computer interfaces. Developers can integrate this API to enable Claude to translate prompts (e.g., “find me a hotel in Colorado”) into specific computer commands.