Preparing a clear path to every model.
Built for developers and teams in Malaysia. Access eligible text, reasoning, image, audio, and video models through one managed Origami API—with hard budgets and request-level usage control.
GATEWAY / JOURNALkey ok72% remainingroute 02route ready1,204 tok$0.011Real requests, metered in real time—this is what your ledger looks like.
Whether you are hacking on a weekend project or governing AI spend across a company, the same key fits both doors.
From idea to first request in minutes. Swap models by changing one string and keep your favorite tools exactly as they are.
One control plane for every model call your company makes, with the guardrails finance and security actually ask for.
Issue scoped virtual keys and set hard limits while OrigamiAI manages credentials, capacity, and pricing behind one API.
Point your existing tools at our endpoint and every model on the platform is suddenly one string away.
import os
import requests
response = requests.post(
"https://api.origamiai.my/v1/chat/completions",
headers={"Authorization": f"Bearer {os.environ['ORIGAMI_KEY']}"},
json={
"model": "origami/model",
"messages": [{"role": "user", "content": "Fold me a demo."}],
},
)Plain answers for developers comparing AI APIs, gateways, and multimodal platforms.
Origami is a managed AI gateway for developers and teams. It provides one account, virtual API keys, and compatible API endpoints for using eligible text, reasoning, image, audio, and video models while Origami operates provider access, routing, budgets, and usage accounting.
Multi-model means an application can use more than one AI model through the same platform. Multimodal means a model or workflow can work with more than one kind of content, such as text, images, audio, or video. Origami supports both patterns through capability-aware model access.
Yes. Origami is a .my managed service built for developers, startups, and product teams in Malaysia and Southeast Asia. It centralizes model access and usage controls behind one API. Origami does not claim Malaysian data residency; teams with location-specific processing requirements should confirm them before production use.
Yes. Origami provides OpenAI-compatible endpoints for chat completions and other supported modalities, plus an Anthropic Messages-compatible endpoint. Existing clients can usually connect by changing the base URL, API key, and model ID.
Origami uses virtual-key and organization budgets with atomic pre-request reservations. Eligible requests reserve a conservative maximum cost before provider work starts, then settle against reported or safely estimated usage in a request-level ledger.
Availability is shown in Origami's live model catalog and can vary by key policy. The API surface supports eligible text, reasoning, vision, image generation, audio, video, embeddings, reranking, moderation, files, and batch workflows. The live catalog is the source of truth for a specific key.
Text, image, audio, and video behind one API. Pick a model, send a request, ship the result—the gateway handles everything in between.