AI startup Anthropic announced on Monday the launch of the first hybrid reasoning model, Claude 3.7 Sonnet, alongside Claude Code, a new command-line tool designed to enhance AI-assisted coding. The new AI model can produce near-instant responses or extended, step-by-step thinking that is visible to the user. "Claude 3.7 Sonnet shows particularly strong improvements in coding and front-end web development," the company said in a blog post on Monday.
Also Read: Anthropic Partners With UK Government to Explore How AI Can Enhance Public Services
Pricing and Availability of Claude 3.7 Sonnet
The Amazon and Google-backed startup said the Claude 3.7 Sonnet model is its most advanced and will be available on all Claude plans, including Free, Pro, Team, and Enterprise. It is also available through the Anthropic API, Amazon Bedrock, and Google Cloud's Vertex AI.
One of the key advancements in this release is the ability for API users to control the "thinking budget", setting a limit on the number of tokens the model can use for reasoning. "This allows you to trade off speed (and cost) for quality of answer," Anthropic said.
"With the new Claude 3.7 Sonnet, users can toggle 'extended thinking mode' on or off, directing the model to think more deeply about trickier questions," the company said in a separate blog post highlighting Claude's extended thinking capability.
In both standard and extended thinking modes, Claude 3.7 Sonnet maintains the same pricing as its predecessor at USD 3 per million input tokens and USD 15 per million output tokens, including thinking tokens.
Furthermore, the company said the Claude 3.7 Sonnet model outperformed competitors in real-world coding tasks, with companies such as Cursor, Cognition, Vercel, Replit, and Canva reporting its ability to generate production-ready code with minimal errors.
Also Read: Anthropic Unveils New AI Model with Computer Use Capability
Claude Code: AI-Driven Coding Assistant
In addition to the model upgrade, Anthropic introduced Claude Code, a new AI-driven coding assistant available as a limited research preview. This tool enables developers to search and read code, edit files, write and run tests, commit and push code to GitHub, and execute command-line operations. Early testing suggests that Claude Code can complete tasks that would typically take developers over 45 minutes of manual work, reducing development time and overhead.
"In the coming weeks, we plan to continually improve it based on our usage: enhancing tool call reliability, adding support for long-running commands, improved in-app rendering, and expanding Claude's own understanding of its capabilities," the AI startup said.
Security and Safety
The release also includes notable improvements in safety and instruction-following. Claude 3.7 Sonnet reduces unnecessary refusals by 45 percent compared to its predecessor while demonstrating better discernment between harmful and benign requests. Anthropic said it has worked with external experts to evaluate security and reliability, addressing challenges such as prompt injection attacks and model transparency.
Also Read: Anthropic, Palantir, and AWS Partner to Bring Claude AI Models to US Defense Operations
Anthropic's Funding
In another development, Anthropic is finalizing a USD 3.5 billion funding round that would value the company at USD 61.5 billion, Reuters reported on Monday, citing two sources familiar with the matter. Anthropic initially aimed to raise USD 2 billion but was reportedly able to increase that amount during talks with investors.
