documentation
Platform
The ZAUTHX402 platform is a unified command center for AI-powered research and x402 endpoint testing. Run up to four different AI models concurrently, monitor their tool calls in real-time, and synthesize findings into actionable reports.
Multi-Model Research
Run 1 to 4 AI models simultaneously on the same research query. The platform supports ChatGPT, DeepSeek, Grok, and Claude, each running independently with their own reasoning chains. Submit a single prompt and watch all four models work in parallel, streaming their findings to individual terminals in real-time.
This multi-model approach isn't just about speed. Different models have different strengths and biases. When multiple models reach the same conclusion independently, confidence increases. When they diverge, that disagreement often reveals edge cases or ambiguities worth investigating. You get multiple perspectives on the same problem without manually switching between tools.
x402 Endpoint Testing
The platform is purpose-built for testing x402 protocol endpoints. Each agent has access to tools that can discover payment requirements, negotiate pricing, and execute USDC payments on Base, all autonomously. Point your agents at an x402-protected API and they'll probe authentication flows, payment validation logic, and response handling.
This is research you can't do with traditional API testing tools. x402 endpoints require valid payments before they respond, which means standard fuzzers and scanners hit a wall at the paywall. Our agents pay their way through, testing what happens after the transaction clears. They can detect issues like improper nonce handling, missing payment verification, or inconsistent pricing across endpoints.
Tool Monitoring
Every tool call made by every agent is logged and displayed in real-time. The timeline view shows HTTP requests, x402 payments, web scrapes, and database operations as they happen. Each entry captures the full request payload, response data, execution status, and timing information.
Click any tool call to expand its details: see the exact JSON sent to an endpoint, the full response body, how long the request took, and whether it succeeded or failed. This visibility is essential for understanding what your agents are actually doing, debugging unexpected behavior, and building confidence that findings are based on real observations rather than hallucinations.
Summarizer
After your agents complete their research, the summarizer synthesizes findings from all four models into a single cohesive report. It identifies areas of agreement, highlights contradictions, and extracts the most important discoveries. The first summary generation is free; retries cost a fraction of a credit.
Summaries are rendered as markdown and can be downloaded directly. This makes it easy to share findings with your team, document research for future reference, or feed results into other tools. The summarizer understands context from the full conversation, so it captures nuances that might be lost if you just skimmed the individual agent outputs.
Chat History
Every research session is saved automatically. Pick up where you left off, review past findings, or continue a conversation with follow-up questions. The platform maintains full conversation history for each agent, so context accumulates across multiple rounds of investigation.
Follow-up queries let you drill deeper with a single agent at reduced cost. If one model found something interesting, you can ask it clarifying questions without re-running all four. Pin important chats for quick access, rename them for organization, or delete them when you're done. Your research library grows with each session.
On this page