Unity ML-Agents

What it does

Unity ML-Agents is a toolkit that turns Unity games and simulations into environments for training intelligent agents. It supports deep reinforcement learning and imitation learning, connecting simulated observations and actions with a learning process.

Developers and researchers use it to experiment with agent behavior in controlled environments. Setup combines the Unity package with compatible Python training tools, and requires defining observations, actions and the task objective. It is an AI-development toolkit rather than the general Unity game engine; training time and the quality of the environment affect results.

What it can do

• Simulation environments
• Reinforcement learning
• Imitation learning
• Agent training
• Training examples

Works with

• Unity

Before you choose

• Unity and model-training infrastructure have their own requirements and terms.
• Simulation performance does not automatically establish real-world reliability.

Listed under

• Coding
• Sandboxes and infrastructure

Get started

1. Check the supported Unity and Python versions.
2. Install the package and training tools.
3. Define the environment, agent observations, actions and training objective.

The toolkit is distributed as open-source software. Unity licensing and the compute used for training are separate.

We couldn’t find published prices for Unity ML-Agents. See their pricing page ↗

Sources for this overview

Product pages, documentation, and pricing checked 30 Oct 2026. Each source supports part of the overview; availability and prices can change.

• GitHub - Unity-Technologies/ml-agents: The Unity Machine Learning Agents Toolkit (ML-Agents) is an open-source project that enables games and simulations to serve as environments for training intelligent agents using deep reinforcement learning and imitation learning. · GitHub ↗github.com

Company website

github.com: product, pricing and policy pages

Social accounts

Accounts the company links, and its posts on X

Human review

Checked by a person

knowledge_base

manual_review

Last checked 30 Oct 2026. Every field keeps its source and date; unknown means we couldn’t confirm it.