Anthropic's Claude AI chatbot has long been one of the best ChatGPT alternatives and now a big update has taken it to another level – including beating OpenAI's GPT-4o model in some industry standard benchmark tests.
Like Google Gemini, Claude is a family of three different AI models. The new Claude 3.5 Sonnet (which takes the baton from Claude 3 Sonnet) is the company's mid-tier AI model, sitting in between the Claude 3 Haiku (for smaller tasks) and the larger 'Opus' model, which is more like GPT-4.
This new Sonnet model now powers the browser-based Claude.ai and the Claude iOS app, both of which you can use right now for free. Like ChatGPT, there are Pro and Team subscriptions available for Claude that let you use it more intensely, but the free version gives you a taste of what it can do.
So what's new in Claude 3.5 Sonnet? The big improvements are its ability to handle vision-based tasks – for example, creating charts or transcribing handwritten notes – with Anthropic calling it "our strongest vision model yet". The company also says that Sonnet "shows a marked improvement in grasping nuance, humor, and complex instructions".
The upgraded Claude is also simply faster and smarter than before, edging out ChatGPT's latest GPT-4o model across many benchmarks, according to Anthropic. That includes setting new benchmark high scores for "graduate-level reasoning", "undergraduate-level knowledge" and "coding proficiency".
This means Claude could be a powerful new sidekick if you need help with creative writing, creating presentations and coding – particularly as it now has a new 'Artifacts' side window to help with refining its creations.
Ultimate homework assistant?
Another handy new feature in Claude 3.5 Sonnet is its so-called 'Artifacts' side window, which lets you see and tweak its visual creations without having to scroll back and forth through the chat.
For example, if you ask it to create a text document, graph or website design, these will appear in a separate window alongside your conversation. You can see an example of that in action in the video above, which shows off Claude's potential for creating graphs and presentations.
So how does this all compare to ChatGPT? One thing Claude doesn't have is a voice or audio powers – it's purely a text-based AI assistant. So if you're looking to chat casually with an AI assistant to brainstorm ideas, then ChatGPT remains the best AI tool around.
But Claude 3.5 Sonnet is undoubtedly a powerful new rival for text-based tasks and coding, edging out GPT-4o in benchmarks and giving us an increasingly well-rounded new option for both creative tasks and coding.
The headline AI battle might be ChatGPT vs Google Gemini vs Meta AI, but if you want a fast, smart AI sidekick to help with a variety of tasks, then it's well worth taking Claude 3.5 Sonnet for a spin in its browser version or iOS app.
You might also like...
- Meta AI vs ChatGPT vs Google Gemini: we tell you which chatbot is the best
- Pro comedians tried using ChatGPT and Google Gemini to write their jokes – these were the hilariously unfunny results
- Ahead of GPT-5 launch, another test shows that people cannot distinguish ChatGPT from a human in a conversation test — is it a watershed moment for AI?
