Breaking: Anthropic Releases Claude Sonnet 4.5 With Revolutionary Coding Capabilities
Claude Sonnet 4.5 is available everywhere today. If you're a developer, simply use claude-sonnet-4-5 via the Claude API. This groundbreaking release marks a significant leap forward in artificial intelligence capabilities, particularly for software development and autonomous agent creation.
Record-Breaking Performance Benchmarks
Claude Sonnet 4.5 hit 77.2 percent on SWE-Bench Verified, jumping to 82.0 percent when given extra computing power. These impressive scores demonstrate substantial improvements over previous models and competing systems. On OSWorld, a benchmark that tests AI models on real-world computer tasks, Sonnet 4.5 now leads at 61.4%. Just four months ago, Sonnet 4 held the lead at 42.2%.
The model represents a paradigm shift in artificial intelligence capabilities. Claude Sonnet 4.5 offers state-of-the-art performance on coding benchmarks. The company says Claude Sonnet 4.5 is capable of building "production-ready" applications, rather than just prototypes, representing a leap in reliability from previous AI models. This evolution from prototype assistance to production-grade development marks a crucial milestone in artificial intelligence advancement.
Revolutionary Autonomous Coding Capabilities
One of the most remarkable features of Claude Sonnet 4.5 is its unprecedented ability to work autonomously for extended periods. Claude Sonnet 4.5 resets our expectations—it handles 30+ hours of autonomous coding, freeing our engineers to tackle months of complex architectural work in dramatically less time while maintaining coherence across massive codebases. This represents a dramatic improvement from previous models, which could only maintain focus for much shorter durations.
Hershey says he's seen Claude Sonnet 4.5 code autonomously for up to 30 hours during early trials with some enterprise customers. In that time, he watched the AI model not only build an application, but also stand up database services, purchase domain names, and perform a SOC 2 audit to make sure the product was secure. These capabilities demonstrate a level of independence and sophistication previously unseen in artificial intelligence systems.
Integration With Major Development Platforms
Claude Sonnet 4.5, Anthropic's most advanced model for coding and real-world agents, is now rolling out in GitHub Copilot to Copilot Pro, Pro+, Business, and Enterprise. The widespread integration ensures developers across various platforms can leverage these advanced capabilities immediately.
Today, we're excited to announce that Claude Sonnet 4.5, powered by Anthropic, is now available in Amazon Bedrock, a fully managed service that offers a choice of high-performing foundation models from leading AI companies. This new model builds upon Claude 4's foundation to achieve state-of-the-art performance in coding and complex agentic applications. The Amazon Bedrock integration provides enterprise customers with robust infrastructure and security features essential for production deployments.
Industry-Leading Safety Features
Claude Sonnet 4.5 is being released under our AI Safety Level 3 (ASL-3) protections, as per our framework that matches model capabilities with appropriate safeguards. These safeguards include filters called classifiers that aim to detect potentially dangerous inputs and outputs—in particular those related to chemical, biological, radiological, and nuclear (CBRN) weapons. This commitment to safety ensures responsible deployment while maintaining powerful capabilities.
The model showcases significant improvements in alignment and safety metrics. Claude's improved capabilities and our extensive safety training have allowed us to substantially improve the model's behavior, reducing concerning behaviors like sycophancy, deception, power-seeking, and the tendency to encourage delusional thinking. These enhancements make Claude Sonnet 4.5 not just more capable but also more reliable and trustworthy for enterprise applications.
Specialized Domain Excellence
Cybersecurity – Claude Sonnet 4.5 can be used to deploy agents that autonomously patch vulnerabilities before exploitation, shifting from reactive detection to proactive defense. Finance – Sonnet 4.5 handles everything from entry-level financial analysis to advanced predictive analysis, helping transform manual audit preparation into intelligent risk management. Research – Sonnet 4.5 can better handle tools, context, and deliver ready-to-go office files to drive expert analysis into final deliverables and actionable insights. These specialized capabilities demonstrate the model's versatility across critical business functions.
Developer Tools and Infrastructure
Alongside the launch of Claude Sonnet 4.5, Anthropic is also launching the Claude Agent SDK. The company says this is the same infrastructure that powers Claude Code and can be used to help developers build their own agents. This release democratizes access to advanced agent development capabilities, enabling developers to create sophisticated autonomous systems.
We're introducing several upgrades to Claude Code: a native VS Code extension, version 2.0 of our terminal interface, and checkpoints for autonomous operation. Powered by Sonnet 4.5, Claude Code now handles longer, more complex development tasks in your terminal and IDE. These enhancements significantly improve the developer experience and productivity potential.
Pricing and Availability
Claude Sonnet 4.5 will be available via the Claude API and in the Claude chatbot. The pricing for developers is the same as Claude Sonnet 4: $3 per million input tokens (roughly 750,000 words, or more than the entire "Lord of the Rings" series) and $15 per million output tokens. This competitive pricing structure maintains affordability while delivering significantly enhanced capabilities.
We recommend upgrading to Claude Sonnet 4.5 for all uses. Whether you're using Claude through our apps, our API, or Claude Code, Sonnet 4.5 is a drop-in replacement that provides much improved performance for the same price. The seamless upgrade path ensures organizations can quickly benefit from the latest advancements without disrupting existing workflows.
Real-World Impact and Customer Success
For Devin, Claude Sonnet 4.5 increased planning performance by 18% and end-to-end eval scores by 12%—the biggest jump we've seen since the release of Claude Sonnet 3.6. It excels at testing its own code, enabling Devin to run longer, handle harder tasks, and deliver production-ready code. These measurable improvements translate directly into enhanced productivity and code quality for development teams.
In a statement shared with TechCrunch, Cursor CEO Michael Truell said Claude Sonnet 4.5 represents state-of-the-art coding performance, specifically on longer horizon tasks. The endorsement from leading development tools underscores the practical value of these advancements for everyday coding tasks.
Future Implications
Claude Sonnet 4.5 represents a fundamental shift in how artificial intelligence can participate in software development and complex problem-solving. "This is a continued evolution on Claude, going from an assistant to more of a collaborator to a full, autonomous agent that's capable of working for extended time horizons," White said. This evolution suggests a future where artificial intelligence becomes an increasingly capable partner in technical and creative endeavors.
The rapid pace of improvement demonstrates the accelerating progress in artificial intelligence capabilities. Anthropic said the rapid progress, marked by major Sonnet updates in February and May, shows a pattern where every six months its new model can handle tasks that are twice as complex. This exponential growth pattern suggests even more transformative capabilities will emerge in the near future.
Frequently Asked Questions
What makes Claude Sonnet 4.5 different from previous models?
Claude Sonnet 4.5 represents a significant advancement in artificial intelligence capabilities, particularly for coding and autonomous agent development. The model can work independently for over 30 hours, compared to just 7 hours for previous versions. It achieves state-of-the-art performance on multiple benchmarks, scoring 77.2 percent on SWE-Bench Verified and 61.4 percent on OSWorld. The model can build production-ready applications rather than just prototypes, marking a crucial shift in practical utility.
How much does Claude Sonnet 4.5 cost to use?
Claude Sonnet 4.5 maintains the same pricing structure as its predecessor, costing $3 per million input tokens and $15 per million output tokens. This pricing applies to API usage, while the model is also available through various subscription tiers on Claude.ai and integrated platforms like GitHub Copilot and Amazon Bedrock. The consistent pricing makes it a drop-in replacement that provides enhanced performance without additional cost.
Which platforms support Claude Sonnet 4.5?
Claude Sonnet 4.5 is available across multiple platforms including Claude.ai (web, iOS, and Android), the Claude API, Amazon Bedrock, Google Cloud Vertex AI, and GitHub Copilot (Pro, Pro+, Business, and Enterprise tiers). The model also integrates with popular development tools like Visual Studio Code, Cursor, and Windsurf. This widespread availability ensures developers can access the model through their preferred platforms and workflows.
What safety measures are included with Claude Sonnet 4.5?
Claude Sonnet 4.5 is released under Anthropic's AI Safety Level 3 (ASL-3) framework, which includes sophisticated safeguards and filters designed to prevent potentially dangerous outputs. The model shows substantial improvements in alignment, reducing concerning behaviors like sycophancy, deception, and power-seeking. It also features enhanced resistance to prompt injection attacks and includes specialized classifiers to detect and prevent misuse related to sensitive topics.
Can Claude Sonnet 4.5 really code autonomously for 30 hours?
Yes, Claude Sonnet 4.5 has demonstrated the ability to work autonomously on complex coding tasks for over 30 hours while maintaining focus and coherence. During enterprise trials, the model has successfully built complete applications, set up database services, purchased domain names, and even performed security audits independently. This represents a dramatic improvement from previous models and enables developers to delegate substantial projects to the artificial intelligence system.
