Anthropic Launches World’s First Hybrid Reasoning AI Model As It Looks To Raise $3.5 Billion

The growing demand for deep thinking generative artificial intelligence (GenAI) technology has resulted in artificial intelligence (AI) companies striving to offer the most capable platforms.

Deep thinking is a feature in which a chatbot thinks long and hard about its answer before generating a response, factoring in a wide range of parameters and criteria and scouring diverse sources from across the web.

This helps users with complex questions that need stronger reasoning and analyses, by thinking for a few seconds to minutes before generating a final response.

In the last month, deep thinking capabilities have been made available through the chatbots of OpenAI, Perplexity, Google, xAI, and others.

However, until now, most AI companies have been offering this capability through independent AI models instead of integrating it into their frontier models—something that Anthropic has opted for with its latest announcement.

Through a news release published on its website, Anthropic announced the release of its newest AI model—Claude 3.7 Sonnet—the world’s first hybrid reasoning model and the startup’s most intelligent model so far.

Claude 3.7 Sonnet brings users the capability of quick responses as well as extended, step-by-step thinking responses—that are made visible to users—in one model.

For this, Anthropic has made its thought process visible in raw form, which enables benefits such as enriched trust, enhanced alignment, and increased interest in watching the model think.

“We’ve developed Claude 3.7 Sonnet with a different philosophy from other reasoning models on the market,” Anthropic said. “Just as humans use a single brain for both quick responses and deep reflection, we believe reasoning should be an integrated capability of frontier models rather than a separate model entirely. This unified approach also creates a more seamless experience for users.”

Essentially, Claude 3.7 Sonnet functions as an ordinary LLM and a reasoning model in one, where users can pick when they want it to answer normally and when to think longer before answering.


Claude 3.7 Sonnet is available on all Claude plans—Free, Pro, Team, and Enterprise—as well as the Anthropic API, Amazon Bedrock, and Google Cloud’s Vertex AI. Its extended thinking capability isn’t available for free users. However, it costs the same as previous models $3 per million input tokens and $15 per million output tokens for both standard and extended thinking modes.

For this tool, Anthropic has chosen to focus more on real-world tasks rather than optimizing for math and computer science competition problems. Additionally, the model shows strong improvements in coding and front-end web development.

Speaking of coding, Anthropic also introduced Claude Code—its first agentic coding tool—which “enables developers to delegate substantial engineering tasks to Claude directly from their terminal.” The tool is currently available as a limited research preview. As such, it can search and read code, edit files, write and run tests, commit and push code to GitHub, and use command line tools.

Furthermore, Anthropic has improved the coding experience on Claude.ai, allowing developers to connect their code repositories directly to Claude.


This announcement comes as Anthropic is reported to be close to concluding a new funding round that would see it raise $3.5 billion at a valuation of $61.5 billion.

This follows previous reports that noted it was initially in talks to raise $2 billion, but the number grew with improved investor confidence.

The upcoming round is expected to witness participation from the likes of Lightspeed Venture Partners, General Catalyst, Bessemer Venture Partners, and MGX (which is also a part of Stargate).

If the new funding hits $3.5 billion, Anthropic would have raised around $18 billion so far.

Do you think Anthropic’s hybrid approach will allow it to capture a stronger position in the GenAI chatbot market? Do you think its competitors should consider making similar moves?

Let us know in the comments below!