Run one prompt through many minds.

Fan out a single prompt across 2–6 proposer models, let one or more judges synthesise the answer, then inspect every per-model response and the fused result side by side. Transparent per-token cost throughout.

1 · PromptMarkdown supported

~1 in tokens0 / 2000

System prompt (optional)

2 · Proposers (3 of 30)2–6 models · fan-out

3 · Judges (1 of 5)synthesis · 1 default

4 · Mode

EU-hosted only

Judge synthesises one merged answer

Complete the quick security check to run

~ $0.00693 proposers + 1 judge(s)Consensus mode · streaming

Live execution

Response will appear here…

Run export

JSON

Download the full run as a JSON file: every proposer and judge call, token counts, latencies, model identifiers and total cost. Useful for record-keeping and review.

Run a prompt to enable export.