Anthropic’s smartest AI is here

Welcome back.

Anthropic just gave Claude a serious upgrade.

With hybrid reasoning and a new coding assistant, Claude 3.7 Sonnet isn’t just faster, it’s smarter, too.

But can it really keep up with the competition? Let’s find out.

In today’s release:

1. Anthropic launches Claude 3.7 Sonnet

2. Alibaba unveils QwQ-Max-Preview

3. Google launches Gemini Code Assist

Claude 3.7 Sonnet launches with hybrid reasoning and Claude Code

Anthropic has unveiled Claude 3.7 Sonnet, its most advanced model yet. It features hybrid reasoning capabilities that allow users to choose between near-instant responses or extended, step-by-step thinking.

Alongside the model, Anthropic also introduced Claude Code, a new agentic coding tool that enables developers to automate complex coding tasks directly from the command line.

• Hybrid reasoning: Users can switch between fast answers or in-depth, reflective reasoning.

• Claude Code preview: A command-line tool that can read, edit, test, and push code, streamlining development.

• Coding powerhouse: Claude 3.7 excels in front-end development, complex refactoring, and full-stack updates.

Claude 3.7 Sonnet is available across all Claude plans, as well as through Anthropic’s API, Amazon Bedrock, and Google Cloud’s Vertex AI. The model retains the same pricing as its predecessors ($ 3/million input tokens, $ 15/million output tokens) and introduces an extended thinking mode, unlocking deeper reasoning for complex tasks.

How can you take advantage of this?

Developers can use Claude Code to automate time-consuming engineering tasks, reducing manual work and boosting productivity. For businesses, the model’s extended reasoning mode can enhance problem-solving in areas like math, physics, and programming, delivering higher-quality results with the same cost-efficiency.

Alibaba unveils QwQ-Max-Preview with advanced reasoning

Alibaba has released QwQ-Max-Preview, the latest addition to its Qwen series. The newest model shows advanced capabilities in deep reasoning, mathematics, coding, and general-domain tasks.

Built on Qwen2.5-Max, this preview offers a glimpse into the full QwQ-Max release while emphasizing Agent-related workflows and logical problem-solving. Alibaba also confirmed plans to open-source both QwQ-Max and Qwen2.5-Max under the Apache 2.0 license, promoting accessibility and community collaboration.

• Reasoning power: Strong in math, coding, and agent workflows.

• Future plans: A Qwen Chat App is on the way, offering AI access to non-technical users.

Alibaba also plans to launch QwQ-32B, a smaller, open-source reasoning model optimized for “privacy-sensitive” and low-latency applications. This lightweight model will allow developers to run AI locally without heavy computational demands, broadening use cases in industries like finance, healthcare, and security.

How can you take advantage of this?

With QwQ-Max-Preview and the upcoming open-source models, developers and businesses can tap into Alibaba’s advanced reasoning capabilities for everything from coding automation to complex problem-solving. Various apps will also benefit from lightweight, local deployment options, giving more flexibility and control over AI projects.

Google launches Gemini Code Assist with high usage limits

Google has announced the public preview of Gemini Code Assist for individuals, offering developers access to AI coding help with generous usage limits.

Powered by Gemini 2.0, the tool supports all major programming languages and offers up to 180,000 code completions per month, far exceeding competitors' limits.

• High limits: 180,000 code completions/month, 90x more than most no-cost coding assistants.

• Wide integration: Available in Visual Studio Code, JetBrains IDEs, Firebase, and Android Studio.

• AI reviews: Gemini Code Assist for GitHub offers AI code reviews for public and private repos..

Gemini Code Assist supports tasks like code generation, debugging, and real-time code reviews. With a 128,000-token context window, developers can input large files for more accurate coding help. The GitHub integration also offers customizable code reviews, allowing teams to tailor AI feedback to their specific style guides.

How can you take advantage of this?

Whether you're a student, freelancer, or startup developer, Gemini Code Assist gives you access to powerful AI tools at no cost. Use it to speed up coding, get detailed code reviews, and tackle complex development tasks without worrying about hitting usage limits or paying for premium plans.

OTHER AI NEWS

Amazon teases major Alexa upgrade: Amazon’s new Alexa AI promises smarter conversations, but most users worry about privacy. A CNET survey shows many won’t pay extra for new features, despite the hype.

Google Drive lets you search for transcripts: Google Drive now lets users search video transcripts directly in the player. The new tool boosts accessibility and challenges paid transcription services like Otter.

Planner 5D simplifies design: Planner 5D uses AI and VR to help users create 2D and 3D floor plans with ease. From hobbyists to pros, it’s a go-to tool for visualizing layouts and home decor.

Apple’s smart home hub leaks: Apple is working on a home hub and a robotic-arm smart display, with launches expected in 2025-2027. The devices will likely feature Apple Intelligence, touch controls, and homeOS.

TRENDING AI TOOLS

1. Zapier Agents: Create your own superhuman teammates in minutes.

2. Carousel Studio: The fastest way to create LinkedIn carousels.

3. Captiwate: Inbound video calls with website visitors.

4. Phedra X: Edit & revisualize images with AI in your browser.

5. TestAI: 1,000+ automated AI tests for AI agents.

AND THAT’S A WRAP

Thank you for reading!

If you found this email useful, share it with a friend or colleague who also loves AI.

Also, drop me a follow on Twitter/X for more AI and tech updates.

See you in the next one!

Mike