Documentation
Product documentation for getting started and building with TensorFusion services.
Choose a product to get started: compute infrastructure, model APIs, or agents.
Compute infrastructure
Tensor OS Enterprise
A full-stack AI compute operating system with GPU pooling, scheduling, monitoring, and operations.
Tensor Engine Open Source
An open-source GPU virtualization engine for partitioning, pooling, and remote sharing, reducing GPU costs by 60–80%.
Model APIs
Agents
AI Gateway / Tensor Gateway (Self-hosted)
The self-hosted edition of Inference Space: the same API on your infrastructure, with multi-model routing, rate limits, and audit logs.
Tensor Box Agent Sandbox
A secure, isolated execution environment for agents with code execution, tool calls, and file operations.
Tensor Agent Platform
An end-to-end platform for multi-model orchestration, workflows, knowledge bases, and monitoring.