Verified AI. One function call. Stop shipping hallucinations.
shield() any LLM output. Every claim fact-checked with evidence chains. Multi-agent pipelines for finance, legal, research, and compliance.
pip install veroqfrom veroq import VeroqClient
client = VeroqClient()
# Ask anything — 41 intents, trade signals, LLM summaries
answer = client.ask("How is NVDA doing?")
# Verify any claim — evidence chains, confidence breakdown
result = client.verify("NVIDIA beat Q4 earnings")
# Multi-agent verified analysis — 5 agents, one verified answer
swarm = client.create_verified_swarm("Analyze NVDA for a long position")
print(swarm["synthesis"]["summary"])| Capability | What you get |
|---|---|
| Verified Swarm | 5-agent pipeline — planner, researcher, verifier, critic, synthesizer |
| Agent Runtime | Domain-specific verticals: finance, legal, research, compliance, custom |
| Verify Any Output | Extract + verify claims from any LLM text — POST /verify/output |
| Agent Memory | Per-agent persistent context — auto-enriches future queries |
| Real-Time Stream | SSE push for ticker status, signal changes, claim updates — GET /watch |
| Verification Receipts | Permanent public receipts with embeddable SVG badges and OG tags |
| Composable Recipes | Marketplace agents run through Verified Swarm with trust scores |
| 62 MCP Tools | Every tool with WHEN TO USE, RETURNS, COST, EXAMPLE |
| Enterprise Safety | Permission engine, decision lineage, escalation, audit trails |
| External MCP | Securely proxy external APIs with trust levels and rate limits |
| Self-Improvement | Feedback loop with web search fallback fills data gaps over time |
| Cost Control | 3 modes (cheap/balanced/premium), per-step budgets, credit transparency |
| Package | Install | |
|---|---|---|
| Python | pip install veroq |
.ask() .verify() .create_verified_swarm() |
| TypeScript | npm i @veroq/sdk |
.ask() .verify() .createVerifiedSwarm() |
| MCP Server | npm i -g veroq-mcp |
62 tools for Claude / Cursor |
| CLI | npm i -g @veroq/cli |
veroq ask veroq verify |
| LangChain | pip install langchain-veroq |
40 tools |
| CrewAI | pip install crewai-veroq |
40 tools |
| Vercel AI | npm i @veroq/ai |
53 tools |
| n8n | npm i n8n-nodes-veroq |
38 operations |
| App | What it does |
|---|---|
| TradingAgents-Pro | 18 AI agents — showcases the verified intelligence layer |
| trading-agent | Autonomous trading — scans, analyzes, verifies, signals |
| financial-research-agent | Multi-step research with evidence chains |
| portfolio-tracker | Real-time SSE portfolio monitoring |
| crypto-monitor | 200 tokens — prices, DeFi, sentiment |
| fact-checker | CLI — verify any claim with evidence |
| veroq-cookbook | 10 ready-to-run examples (Python + TypeScript) |
/ask— 41 intents, LLM summaries, SSE streaming, composite trade signals (0-100)/verify— Evidence chains, confidence breakdown (4 factors), source reliability scores, verification receipts/swarm/run— 5 verified agents coordinate on any query. Every step verified with decision lineage./runtime/run— Domain-specific pipelines with pre-configured safety rules per vertical/verify/output— Extract + verify claims from any LLM text (5cr + 2/claim)/memory/store+/memory/recall— Per-agent persistent context, auto-enriches/ask/watch— Real-time SSE for ticker status, signal changes, confidence updates, claim invalidations/verify/receipt/:id— Permanent public verification receipts with embeddable SVG badges- 1,061+ tickers — equities, crypto, ETFs, commodities + auto-discovery
- 200+ sources across 18 categories
- 300+ API endpoints when you need granular control
veroq.ai · API Reference · Docs · Pricing
Free tier: 1,000 credits/month. No credit card required. Enterprise plans for audit trails, escalation, and external MCP.