On Monday, Anthropic announced the launch of its latest frontier model, Claude Sonnet 4.5, which it says delivers state-of-the-art results on coding benchmarks and a leap in reliability for software engineering tasks. Unlike earlier versions, the company claims this model can autonomously build “production-ready” applications, not just prototypes.
Claude Sonnet 4.5 is available starting today through the Claude API and the Claude chatbot. Pricing remains unchanged from Claude Sonnet 4, at $3 per million input tokens and $15 per million output tokens, with one million tokens equating to roughly 750,000 words.
Anthropic’s models have become favorites among developers and enterprises, with companies like Apple and Meta reportedly using Claude internally. The AI has also been integrated into popular coding tools such as Cursor, Windsurf, and Replit. However, competition is heating up as OpenAI’s GPT-5 has recently outperformed Claude models on several coding tasks.
Anthropic says Claude Sonnet 4.5 is industry-leading on benchmarks like SWE-Bench Verified, but stresses that the model’s true capabilities go beyond test scores. In early trials, the AI was observed autonomously coding for up to 30 hours, building applications, deploying databases, purchasing domains, and even conducting SOC 2 security audits.
“Claude Sonnet 4.5 represents state-of-the-art coding performance, especially on longer horizon tasks,” said Michael Truell, CEO of Cursor. Windsurf CEO Jeff Wang echoed that sentiment, calling it a “new generation of coding models.”
Beyond coding, Claude Sonnet 4.5 is also billed as Anthropic’s most aligned model yet, with reduced tendencies toward sycophancy, deception, and prompt injection vulnerabilities.
The company is also introducing the Claude Agent SDK, the same infrastructure that powers Claude Code, enabling developers to build custom agents. Additionally, Anthropic is releasing a research preview called “Imagine with Claude” for Max subscribers, showcasing real-time software generation with no pre-written code.
Related: Anthropic Faces Brief Outage Affecting APIs, Console, and Claude AI
The launch underscores the frenzied pace of the AI race, with new flagship models arriving in rapid succession. Claude Sonnet 4.5 comes less than two months after Anthropic rolled out Claude Opus 4.1, reflecting the difficulty of holding a lasting competitive edge in the market.
