Artificial intelligence startup Anthropic isn’t resting on its laurels. Mere weeks after releasing Claude Sonnet 3.5 – the latest addition to the Claude family of large language models (LLMs) – the company has just announced a raft of new updates to its Anthropic console.
The latest iteration of Claude has already claimed the top spot in the S&P Global leaderboard of Large Language Models (LLMs), beating out rivals such as OpenAI’s Chat GPT and Google Gemini rivals. Now it’s going even further by providing users with the tools to “generate, test, and evaluate prompts” in the aim of delivering the most efficient outcome for your needs.
Recognizing that generating efficient prompts can be a challenge, but these new features – including the ability to generate test cases – promise to “make it easier for users to produce high quality prompts.”
What New Features Does Claude Sonnet 3.5 offer?
The new and improved Claude announced by Anthropic can assist with generating prompts, conjure up test cases, and lets you compare a range of prompts, making it less of a chore to get the results you want. It’s also available for free on Claude.ai and the Claude iOS app.
You can now let Claude generate a prompt for you based on a simple description outlining your requirements. The new test case generator function produces input variables and allows you to interrogate that prompt to see how well it meets your needs and responds to a range of request.
🔎 Want to browse the web privately? 🌎 Or appear as if you're in another country?
Get a huge 86% off Surfshark with this special tech.co offer.
You also have the option to create a test suite with the “Evaluate” feature, too. This lets users see how their prompts respond to “a range of real-world inputs,” and directly in the Anthropic console – saving you the trouble of toggling between a dozen different documents. You can add tests cases manually or ask Claude to generate some for you. Modify as needed, click “Run all,” and voila! You’re then free to tinker until your prompt provides the ideal outcome.
Anthropic now also allows users to compare the outputs of two or more prompts side by side, which makes it far simpler to evaluate the impact of any changes you’ve made.
Looking for further details on how to generate your own prompts? Check out the extensive user guide from Anthropic.
Take the Pain Out of Writing Prompts
Struggling to refine your prompt writing skills? A well-crafted text can make all the difference in getting the results you’re seeking, as users of Microsoft’s ubiquitous AI companion Copilot have recently lamented. Yet, unlike Microsoft, Anthropic has acknowledged these difficulties and aimed to make its software even more accessible.
So, whether you use Claude, ChatGPT, or Gemini, there are a number of ways to ensure your prompt-writing is getting the results you need:
- Consider taking an AI course like that offered by LinkedIn.
- Explore your options with some of the best AI chatbots available
- Read about how to craft an effective ChatGPT prompt
If you’re looking to get the best results from your chatbot, steer away from subjective language and provide plenty of contextual information. Also, reference materials if you’re looking to get domain-specific knowledge.
