Ground a weave in current sources with web search

Intermediate4 min

Turn on web search so models answer from current sources rather than training memory. Essential for anything recent or fact-sensitive.

Before you start

An LLMWeave account

Enable web search on the weave

When composing the weave, turn on web search. The models will pull from live sources before answering, which grounds the result in current information.

Write a specific, searchable prompt

Web search grounds the model in what is indexed. The more specific your prompt, the better the sources it retrieves. Vague prompts return vague grounding.

Run and check the grounding

Run the weave and review how the answer uses the retrieved sources. For comparisons or research, this is the difference between a current answer and a stale one.

Keep going

Beginner · 4 min

Compare Claude, GPT, and Gemini in one run

Send a single prompt to several frontier models at once and see their answers together, so you can judge accuracy, tone, and cost in one place.

Intermediate · 7 min

Set up a multi-step workflow in The Loom

Chain multiple steps into a durable workflow that runs in the background and can pause for human review. This is where LLMWeave goes past a single prompt.

Try it yourself

Put this pattern to work on your own task.

Create a weave