We are live
Today, AskVerdict is open to the public.
AskVerdict is a multi-agent AI debate engine. You bring a decision or question — AskVerdict deploys specialized AI agents that argue both sides, challenge each other's assumptions, and return a structured verdict with confidence scoring, evidence citations, and actionable recommendations.
What AskVerdict does
Every decision has competing considerations. Most tools give you one answer. AskVerdict makes the tension visible.
When you submit a question, the engine:
- Deploys advocacy and critique agents that argue opposing positions
- Runs a cross-examination pass where agents directly challenge each other's reasoning
- Surfaces logical fallacies and weak assumptions in real time
- Generates a structured verdict with a recommendation, confidence score, and evidence citations
The output is a complete, auditable record of the decision process — not just a conclusion.
Debate modes
Three inference modes balance speed and depth:
- Fast — Directional analysis in seconds. One credit. Best for quick checks.
- Balanced — Thorough multi-round debate. Three credits. The default for most decisions.
- Thorough — Maximum depth with extended agent reasoning. Eight credits. Best for high-stakes choices.
Verdict Compare
Verdict Compare evaluates 2–5 options side-by-side in a single structured run.
Instead of running separate debates, the engine analyzes all options against shared criteria, surfaces trade-offs across dimensions, and returns a ranked recommendation with supporting rationale.
Useful for: vendor selection, hiring decisions, strategy choices, product direction, technology evaluation.
Decision intelligence tools
The verdict is the starting point. Beyond it:
- Debate Analysis — Post-verdict tab with argument quality graphs, logical fallacy detection, evidence scoring, and reasoning quality metrics
- Debate Assistant — Persistent AI chat anchored to your debate context — ask follow-ups, explore edge cases, request deeper analysis
- Verdict Monitor — Track a decision over time and receive alerts when the underlying recommendation shifts
- Verdict Analyze — Upload documents and run AI-powered analysis against your own content
- Global Chat — Session-persistent assistant with access to your full debate history
Bring your own API keys
If you use Anthropic (Claude), Google Gemini, DeepSeek, Groq, or OpenRouter, you can supply your own API keys and run debates without consuming platform credits.
BYOK keys are stored in your browser only — never transmitted to our servers.
Pricing
AskVerdict AI is free to start. 10 credits on signup, all modes unlocked, no card required.
Paid plans add monthly credits, Verdict Compare, and advanced tools.
| Plan | Price | Credits |
|---|---|---|
| Free | $0 | 10 credits on signup |
| Pro | $29/mo | 500 credits/month |
| BYOK | $19/mo | Unlimited with own keys |
| BYOK Pro | $39/mo | Unlimited + all features |
What comes next
Immediate roadmap priorities:
- Team workspaces and collaborative decision programs
- Expanded debate assistant capabilities and context depth
- Broader model routing and additional provider support
- Enterprise governance controls
Updates post here as features ship. Subscribe to the RSS feed to stay current.
Thank you to every beta user who gave feedback during development. The debates you ran shaped what shipped today.
- The AskVerdict Team