Running work
Batches, the live monitor, reading results, and not paying twice for the same step.
9 articles
- Running a batchOne press, every item at once across all your keys — with two switches that decide whether it storms ahead or takes its time.
- Documents as inputs: PDF, Word, ExcelGive each item a document — a PDF spec sheet, a Word brief, an Excel price list — and a Claude or Gemini step reads it.
- Re-run from a stepChange one step and run it again without paying for — or re-rolling — everything before it.
- Run a spreadsheet of itemsCopy your rows from Google Sheets or Excel, paste them in, press Run, and take the results back out as a spreadsheet. Or upload a CSV.
- The run monitor: watching your keys workA lane per API key showing how the batch was spread, live meters against each key's limits, and a one-click re-run of only what failed.
- Reading and reviewing resultsText as formatted text, images as a grid with a full-size viewer — keep the good ones, regenerate only the rest, and drill into what each step produced along the way.
- Reuse a step's output instead of paying twiceTick “Reuse output” on an expensive step and an identical input serves the cached result at $0 — so you can iterate on what comes after it.
- What a run costs (and what the number means)An estimate of what your provider will bill you, booked per item and per model — visible before you commit, and never a charge from FlowArray.
- Free test mode (mock runs)Any step without a real key pool runs against a built-in mock at $0 — perfect for proving a flow's shape, useless for judging its output.