Getting started
Getting started
A weave sends your task to several models at once and returns one answer. This quickstart walks you from sign-up to your first run, then points you at the tutorials for the patterns worth learning next.
1. Create your account
2 minSign up with your email. The free plan needs no card and includes hourly runs on the free model set, so you can try real weaves before deciding anything.
2. Start a new weave
1 minFrom your dashboard, click New Weave. A weave is one task run across several models at once. You write the prompt once, and LLMWeave sends it to each model and brings the answers back together.
3. Start from a template, or from scratch
2 minTemplates are ready-made recipes for working with several models (consensus draft, rank and fuse, best of N, debate and decide). Pick one to see how a good weave fits together, or start from scratch and choose the models yourself.
4. Run it and read the synthesis
2 to 5 minRun the weave. You'll see the per-model work and then a single synthesized answer that merges the strongest parts. The estimate shows cost before you commit; the final cost is tracked per model.
5. Chain, share, or export
1 minContinue the conversation to refine the result, share a public link, or export the output. From here, the tutorials cover the patterns that make multi-model worth it.
Next: learn the patterns
Compare Claude, GPT, and Gemini in one run
Put Claude, GPT, and Gemini on the same prompt and read the answers side by side. It is the fastest way to catch tone and accuracy differences before you pick a final answer.
BeginnerBuild a consensus draft that merges multiple models
Use the Consensus Draft pattern when you want one clean answer, not a pile of model replies to sort through by hand.
IntermediateRank and fuse multiple model outputs
Generate a few candidate answers, rank them, then fuse the best parts. Use this when quality matters more than speed.
IntermediateGround a weave in current sources with web search
Turn on web search when the answer depends on today’s facts: prices, product pages, news, docs, or anything else that can go stale.
IntermediateSet up a multi-step workflow in The Loom
Use The Loom when one prompt is not enough. Build draft, critique, research, and review steps, then let the workflow run in the background.
IntermediateRun a workflow from the API
Call a saved workflow from code, Zapier, Make, or n8n. Start with polling, then switch to webhooks when the run needs to feed another system.