Chain builder

Build a multi-lens analysis chain

Choose a few lenses for the current page or text, then combine their analyses into one useful answer.

Simple mode keeps every lens on the original source.

Chain details

Profile image

PNG, JPG, or WebP. Max 5 MB or 2000×2000 px.

Tags

Search existing tags or type a new one. Up to 12 tags.

    LensesSteps

    Each lens analyzes the same source. Add up to six, reorder them if helpful, and SoWhatify will combine their results. New to this? The chain guide walks through it with examples.

    Steps can be lenses, raw prompt pairs, other chains, fetched URLs or fixed reference text, and each can choose its own AI model. Use Change InData to choose upstream inputs, and reference outputs in prompts with {$Step name} when a step has a unique name. The chain guide covers every step type, conditions, loops and limits.

    Chain seed input

    Seed: Seed data (Web page, pdf upload, clipboard or text input)

    Available to every step when no upstream sources are selected via Change InData.

    No lenses added yet Choose two or three lenses to create your first chain.
    1. Step 1: Final answer: Final aggregation

      InData

      Add lenses above first

      Final answerFinal aggregationRequired

      Choose how the lens analyses should be combined.

      This prompt receives all refined step outputs and produces the final combined answer. Reference a step output by unique name with {$Step name}, by stable step ID with {$step_…}, or by position with {$step_number_4} (step 4). Step-number references are sticky: if you insert or reorder steps, the number in the user prompt updates to follow the same step.

      Optional. If empty, this step uses the widget-selected model before the chain run model.

      Optional. If empty, the default final synthesis system prompt is used.

      Add a lens to see the number of AI calls.