Orion, the visual agent that acts.

Understand, reason over, and act on images, video, and documents through a single OpenAI-compatible API.

Try Orion for Free Read the Whitepaper

Loved by leading

AI companies

Interact with images, videos, documents in a single API.

Frontier models like GPT, Claude, and Gemini can describe what they see – but they can't act on it. Orion unites the reasoning power of large Vision-Language Models with the accuracy of specialized computer-vision tools – all through one unified API.

One interface for all your visual AI needs.

Images, documents, and videos – all through a single chat-completions interface.

Compose through conversation.

Chain visual operations like: detect → crop → enhance → analyze in a single conversation.

Integrate at warp speed

Drop-in replacement for the OpenAI SDK. Same API pattern – new visual powers.

Auditable outputs

Every response comes with visual proof. Build and integrate with confidence.

Ridiculously versatile.

Whatever your visual task – Orion knows how to act. Built with dozens of specialized computer-vision and multi-modal tools.

Image Intelligence

Document Intelligence

Video Intelligence

123456789101112131415161718

from openai import OpenAI

client = OpenAI(
  base_url="https://api.vlm.run/v1/openai",
  api_key="<VLMRUN_API_KEY>"
)

result = client.chat.completions.create(
  model="vlmrun-orion-1:auto",
  messages=[{
    "role": "user",
    "content": [
      {"type": "text", "text": "Parse this invoice into JSON."},
      {"type": "image_url", "image_url": {"url": "https://.../invoice.pdf"}}
    ]
  }],
  response_format={"type": "json_object"}
)

For Developers

Designed for developers.

Familiar API, unfamiliar power.

All your favorite vision tools, in a single box.
Drop-in replacement for OpenAI SDK.
Handles images, documents, videos via URL or upload.
Streaming support for real-time responses.
Structured outputs with Pydantic / Zod support.

API Docs

For Enterprises

The new visual intelligence layer for your enterprise.

Deploy securely inside your VPC or private cloud – bringing visual intelligence directly to your infrastructure. Power document, image, and video understanding across teams. SOC 2 Type II and HIPAA-ready.

Orion FAQs

Orion is our flagship visual agent. It plans a visual task, calls the right computer-vision and multi-modal tools, validates what comes back, and returns a structured answer. One agent covers images, video, and documents through the same chat-completions interface.

Try Orion Free today.

Chat with Orion Book a Demo