JigsawStack Logo

Beta

Prompt Engine now with Internet, Streaming, Optimization and more!

Share this article

Prompt Engine now with Internet, Streaming, Optimization and more!

Hey devs! 👋

Imagine running any prompt and getting the optimal model to execute it without ever worrying about picking LLMs, managing GPUs, tokens, or rate limits. Now, with the new features, the Prompt Engine is more powerful and flexible than ever.

What’s New?

Prompt Streaming

Want real-time responses? With streaming, results start flowing as the prompt is processed. This is perfect for live apps like chatbots or realtime dashboards.

Prompt Template Page

Not sure where to start or what prompt to write? Prompt templates is a gallery of prompts that are pre-optimized and ready to run with the prompt engine. Whether it’s summarizing articles or extracting key data, you can now find pre-built prompts that are optimized to deliver the best results.

Internet Access

Need up-to-the-minute information? The Prompt Engine now has access to AI search allowing it fetch live data from the web.

Prompt Optimization

We’ve made your life easier. The engine now automatically optimizes prompts to improve accuracy, minimize token usage, and reduce costs. No more tweaking and testing—just plug and play.

Let’s Build Something!

Step 1: Create a free JigsawStack account.
Step 2: Grab your API key from the dashboard.
Step 3: Install the SDK and start building!

Prompt Streaming and Internet Access

Execute Directly for Quick Results

For one-time prompts, skip the setup and run directly. Let’s see Prompt Engine in action— first a run without internet access:

Output

Now let’s enable internet access for live results:

Output

How Fast Did It Run?

Speed matters, especially for real-time applications. Here’s how quickly the Prompt Engine executed with streaming on:

  • Without Internet Access: Response delivered in ~1.41 seconds.

  • With Internet Access: Response streamed live in ~3 seconds, including fetching and formatting up-to-date web data.

The combination of low latency and live streaming ensures you get results fast âš¡

Prompt Optimization

The prompt engine takes the hassle out of fine-tuning by automatically optimizing your prompts for better results. Here's a simple example of how it works:

Output

By automating the fine-tuning process, you can now focus on the outcome rather than the intricacies of crafting the perfect prompt. The engine enhances accuracy, reduces token usage, and minimizes costs— All while ensuring your prompts are tailored for maximum effectiveness.

Just plug in your ideas, and let the engine do the heavy lifting empowering you to streamline your workflows and achieve more, faster!

Prompt Templates

Lets try running one of our prompts from the prompt template page:

Here’s our prompt:

Run the prompt id:

Output

Why We Are Excited!

With these new features, JigsawStack's Prompt Engine is your go-to tool for seamless, real-time AI-powered workflows. Whether you’re optimizing processes, building chatbots, or generating data-driven insights, the Prompt Engine empowers you to deliver smarter solutions at scale.

👥 Join the JigsawStack Community

Have questions or want to show off what you’ve built? Join the JigsawStack developer community on Discord and X/Twitter. Let’s build something amazing together!

Share this article