Build with
sovereign AI.
The PolpAI API gives you 150 specialized agents, 410+ endpoints, and zero data extraction. While Limova caps at 9 agents and ChatGPT offers 1 generic model, PolpAI delivers 150 purpose-built agents. Ship faster. Stay sovereign.
# pip install polpai import polpai # Initialize client client = polpai.Client( api_key="pk_live_..." ) # Run Deal Hunter agent result = client.agents.execute( agent="deal_hunter", task="Find MacBook Pro M4 deals", context={ "budget": 2500, "currency": "CAD" } ) print(result.recommendation) # {price: 2249, savings: 251}
410+ endpoints. 8 tentacles.
Every agent is accessible via REST. JWT authenticated. Rate-limited per plan. Zero data retention after response.
/agents/executeRun any of 150 agents by name. Pass context, task, and optional parameters. Returns structured result + confidence score.
/agents/listList all available agents grouped by tentacle (T1–T8). Filter by domain, capability, or patent family.
/memory/storeWrite to sovereign memory layer (L1–L5). AES-256-GCM encrypted. Stored on your infrastructure via BYOS.
/memory/recallRecall from memory with semantic search. Supports cross-session context, decay scoring, L5 relational anchors.
/deals/scanTrigger Deal Hunter (T1). Returns price data, discount authenticity score (0–1), retailer comparison table.
/gifts/recommendContext-aware gift recommendations. Pass recipient profile, occasion, budget. Returns ranked suggestions with reasoning.
/pme/auditShadow AI compliance audit. Detects unauthorized AI usage, quantifies risk, auto-generates proposal. Value: $1,500+.
/quantum/auditPQC security assessment (T5). Scans infrastructure for post-quantum readiness. Report value: $1,500–$3,000.
/finance/cashflowCash flow analysis and forecasting (T4). Multi-currency, FX risk, invoice automation, fiscal planning.
/chatConversational interface to the full agent team. Session-aware. Supports natural language agent routing.
/agents/metricsPerformance metrics: confidence score, execution time, fitness evolution, predict()/evolve() trajectory.
/healthSystem status, agent availability by tentacle, current API version. No authentication required.
Scale with your usage.
No per-seat pricing. Pay for API calls. Upgrade or downgrade anytime.
- All tentacles + priority
- Email support
- 300 req/min
- Webhook events
- All agents + BYOS
- Priority support + SLA
- 1,000 req/min
- Dedicated endpoints
- Self-hosted deployment
- Custom agent cluster
- Custom SLA + MSA
- PME pipeline included
Your language. Our agents.
Official SDKs for the most popular languages. Community SDKs for the rest.
Python
Full async support. Pydantic models. FastAPI integration.
JavaScript / TypeScript
Node.js + browser. React hooks. Zod validation.
Rust
Zero-overhead bindings. Tokio async. WASM-ready.
OWASP compliant. Sovereign by default.
Every API call protected at the structural level — not just policy level. 56 patent families and 682 unique patent IDs covering the security architecture.
JWT Authentication
Short-lived tokens with automatic rotation. No static API keys in production.
OWASP Top 10
Injection scan on every input. ASVS Level 2. Automated threat modeling per release.
Rate Limiting
Per-key, per-IP, per-agent rate limits. DDoS protection via Cloudflare edge.
Zero Data Retention
API responses are ephemeral. No conversation stored server-side unless explicitly opted in.
Adversarial Compliance
95%+ ACT score across all agents. 50-scenario adversarial test suite per release.
Key Sovereignty (B27)
AES-256-GCM key vault. GitHub PAT rotation. Cloudflare token management. Zero-chat transmission.
PolpAI ecosystem