Claude 3.7 Sonnet + Gemini Thinking

Claude 3.7 Sonnet just dropped!

It's looking like a big week in AI with numerous announcements expected from the big players. It's only Monday and Anthropic just dropped their long awaited new Claude model. The company is calling Claude 3.7 Sonnet their most intelligent model to date and say that they've been focused on real-world tasks instead of the maths and computer science focus of their previous generation models. We've been seeing a lot of "reasoning models" recently and Anthropic are taking a slightly different approach here, users can specify how long Claude should "think" for, allowing it to function like a traditional LLM (with zero thinking time) or like one of the newer reasoning models.

Gemini "Thinking"

Claude 3.7 Sonnet isn't the only new model that's been added to AnyModel today. Google has now joined the fray with their own reasoning model, Gemini Flash 2.0 Thinking. The model is still experimental for now, so you may experience some odd behaviour - but you can now compare Gemini Thinking to the other reasoning models available on AnyModel.
­

Qwen Family

There's been a lot of noise around DeepSeek in recent times, but it's not the only top-tier AI model hailing from China. We're excited to announce that Alibaba's Qwen model family is now available on AnyModel. This includes their flagship, Qwen Max which can beat DeepSeek in a variety of benchmarks. Try it out for yourself on AnyModel!

More models?

I know we've already covered a lot today, but you'll also find 2 new variants of Mistral available on AnyModel (Nemo & Tiny). Today we also announce the addition of MiniMax Text, Magnum v4 and Jamba 1.5 mini. As I alluded to at the top, this is expected to be a big week in AI, and we may have more announcements for you later in the week!