https://api.piramyd.cloud/v1.
Key Capabilities
Chat Completions
Fully OpenAI-compatible
POST /v1/chat/completions endpoint. Drop in Piramyd’s base URL with your existing OpenAI SDK and start routing to any supported model immediately.Anthropic Messages
Native Anthropic-compatible
POST /v1/messages endpoint. Use x-api-key or Authorization: Bearer — your existing Anthropic client works without code changes.System One Decisions
Typed, calibrated decisions via
POST /v1/systemone. Submit a state and a map of noul, choice, or score questions and receive structured probabilistic answers in under 500 ms.Web Search (SIRS)
Multi-engine web search via
POST /v1/search. Powered by SIRS (SearXNG), with optional per-result full-page content extraction in the same request.Web Scraping & Crawling (CROWD)
Single-URL scraping, full site crawls, URL mapping, and structured LLM extraction via
POST /v1/scrape, /v1/crawl, /v1/map, and /v1/extract. Powered by CROWD (Firecrawl).Context Compaction
Long conversations that exceed a model’s context window are automatically compacted. Pass a
thread_id in any request and the API handles the rest transparently.Tool Calling
Function calling with automatic argument buffering, JSON repair, and mid-stream continuation. Works with any model that reports
supports_tools: true.Vision
Send images alongside text in any chat or responses request. Works with any model that reports
supports_vision: true. Upload images via POST /v1/images/upload to avoid base64 payload limits.Base URLs
Use the/v1 prefix for all production integrations. The internal services (search.piramyd.cloud, crawl.piramyd.cloud) are proxied through the main API — you never call them directly.
Plans & Rate Limits
Piramyd has no free forever tier. You start with a 3-day trial (card required) on the Solo plan, and then choose a paid plan to continue. Tokens are unlimited on all paid plans — throughput is governed by requests per minute (RPM).
Additional RPM (+5 RPM) is available as a paid add-on on eligible plans. Manage your plan and purchase add-ons from the dashboard.
A
402 response means you’ve hit a billing limit. A 403 response means a feature or model requires a higher tier. Neither indicates a weekly token budget — tokens are unlimited on paid plans.Next Steps
Quickstart
Make your first API call in under five minutes using Python or JavaScript.
Authentication
Learn how to obtain and use API keys and JWTs.