> For the complete documentation index, see [llms.txt](https://docs.hex-rays.com/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.hex-rays.com/editions-and-licenses/license-types/ida-pro-headless/overview.md).

# Overview

## What is IDA Pro Headless?

IDA Pro Headless is IDA's analysis engine without a user interface. It runs [idalib](https://docs.hex-rays.com/core/idalib/overview.md), IDAPython, the [Domain API](https://docs.hex-rays.com/developer/domain-api.md) and `idat` batch mode, so your scripts and AI agents can open binaries, analyze them, decompile them and query the results with the same engine your analysts use in IDA Pro.

It ships as one ready-to-use package per platform, and it is licensed through a [Hex-Rays license server](https://docs.hex-rays.com/editions-and-licenses/license-server/admin/headless-license-server-setup.md), with **headless seats**: seats taken by scripts, pipelines and AI agents rather than by people.

## What is it made for?

IDA Pro Headless is made for analysts and their scripts and agents working together in the Hex-Rays ecosystem. Typical uses are:

### Automated binary analysis

A script calls IDA from your internal pipelines, without any agent involved: malware triage queues, analysis on every build in CI/CD, diffing binaries against prior releases, or research across firmware archives and sample collections.

### Agentic workflows

Several AI agents analyze binaries at the same time, each with its own seat. Your coding agent drives IDA directly through idalib or through the [IDA MCP Server](https://docs.hex-rays.com/add-ons/mcp/overview.md), while your analysts keep working in IDA Pro on the same license server.

### Integration into a platform you operate

IDA runs behind a service or platform that you host and operate, on your premises or as a SaaS offering, for example to provide disassembly or decompilation to your own users. The license server runs on your side, exactly as for the other uses.

{% hint style="info" %}
For any other distribution model, [contact us](https://hex-rays.com/contact) to scope licensing and deployment.
{% endhint %}

## IDA Pro and IDA Pro Headless

|                     | IDA Pro                                             | IDA Pro Headless                                                                                                                                                               |
| ------------------- | --------------------------------------------------- | ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------ |
| Analysis engine     | Disassembler, decompilers, IDAPython and Domain API | The same engine, without the GUI                                                                                                                                               |
| Made for            | Analysts at their desk                              | Scripts, pipelines, agents and the platforms you operate                                                                                                                       |
| License assigned to | A person, a machine or a floating seat              | A headless seat on your license server, taken while a script or agent runs                                                                                                     |
| Decompilers         | Depends on the edition                              | All local decompilers, see [Decompiler Availability by Edition](https://docs.hex-rays.com/editions-and-licenses/ida-features-by-edition/decompiler-availability-by-edition.md) |

Without the GUI, IDA starts faster, uses less memory, and more workers fit on the same host.

## Key features

### Seats used only while a script or agent runs

A script or agent takes a headless seat when it opens a database and gives it back when it closes it. Ten scripts in sequence need one seat; ten scripts at the same time need ten. See [Headless seats](https://docs.hex-rays.com/editions-and-licenses/license-types/ida-pro-headless/concepts/headless-seats.md).

### Visibility on your usage

The license server shows which machines and instances hold seats, so you can size your pool of headless seats from real usage and scale it from there.

### Offline and air-gapped machines

A machine that cannot reach the license server works with a seat borrowed for it, for a fixed period. See [Work offline with a borrowed seat](https://docs.hex-rays.com/editions-and-licenses/license-types/ida-pro-headless/how-tos/work-offline.md).

### Ready for agents

Each package includes an `AGENTS.md` file that tells a coding agent how to set up the environment, use the idalib API and handle seats. See [Use IDA Pro Headless with coding agents](https://docs.hex-rays.com/editions-and-licenses/license-types/ida-pro-headless/how-tos/use-with-coding-agents.md).

### Docker

On Linux and Windows x64, the package includes what you need to build a container image. See [Run IDA Pro Headless in Docker](https://docs.hex-rays.com/editions-and-licenses/license-types/ida-pro-headless/how-tos/run-in-docker.md).

### Plugins

IDA Pro Headless ships with the [plugins of IDA Pro](https://docs.hex-rays.com/add-ons/plugins/concepts/plugins-shipped-with-ida.md) that work without a user interface. Plugins that only work in the GUI are not included.

## License

### Human seats and headless seats

Your license server can serve two kinds of seats side by side:

* **Human seats**, for analysts working in IDA Pro.
* **Headless seats**, for everything that is not human: scripts, pipelines, AI agents and services.

Both work the same way. A seat is taken while IDA works on a database, and goes back to the pool as soon as the database is closed. A script or agent that is not running holds no seat, so you pay for how many scripts and agents work **at the same time**, not for how many scripts or binaries you have.

### An example

A team of two analysts runs a malware triage pipeline and two AI agents, on a license server with **2 human seats** and **3 headless seats**:

| Time  | Analysts in IDA Pro | Scripts and agents running                     | Human seats in use | Headless seats in use |
| ----- | ------------------- | ---------------------------------------------- | :----------------: | :-------------------: |
| 09:00 | Alice               | Triage script, one sample after the other      |       1 of 2       |         1 of 3        |
| 11:00 | Alice, Bob          | Triage script, and 2 agents on one binary each |       2 of 2       |         3 of 3        |
| 13:00 | Bob                 | 2 agents                                       |       1 of 2       |         2 of 3        |
| 22:00 |                     | Nightly batch of 500 samples, 3 at a time      |       0 of 2       |         3 of 3        |

At 11:00, every headless seat is in use: a fourth script waits for a free seat, or fails. See [Wait for a free seat](https://docs.hex-rays.com/editions-and-licenses/license-types/ida-pro-headless/how-tos/wait-for-a-free-seat.md). The license server console shows these peaks, so you can size your pool from real usage.

{% hint style="success" %}
Start small and scale bottom up: take a few headless seats, watch your peak usage, and add seats when your scripts start to wait.
{% endhint %}

See [Headless seats](https://docs.hex-rays.com/editions-and-licenses/license-types/ida-pro-headless/concepts/headless-seats.md) for the details: when a seat is taken, naming seats, and borrowed seats.

### What your agreement covers

Using IDA with headless seats requires a **sales order that is an addendum to the IDA license agreement (EULA)**. It states:

* the use case it authorizes: automated analysis, agentic workflows, or a platform you operate,
* the number of **concurrent headless seats**.

Each headless seat gives access to the full analysis engine, with all local decompilers. See [Decompiler Availability by Edition](https://docs.hex-rays.com/editions-and-licenses/ida-features-by-edition/decompiler-availability-by-edition.md).

### What you need to run it

* A Hex-Rays license server, version 9.5.2 or later, that serves your headless seats. If your analysts already use floating licenses, upgrade that server and add your IDA Pro Headless license to it: you do not need a second one. See [License server setup](https://docs.hex-rays.com/editions-and-licenses/license-server/admin/headless-license-server-setup.md).
* No license file on your machines: each script or agent finds the server through the `IDA_LICENSE` environment variable. The server does not tie seats to a machine or check MAC addresses.
* For machines that cannot reach the server, [borrowed seats](https://docs.hex-rays.com/editions-and-licenses/license-types/ida-pro-headless/how-tos/work-offline.md).

### Already running scripts with IDA Pro?

The standard IDA license agreement covers IDA Pro used by a person, on a named computer or with a floating seat. Running scripts, pipelines or agents on those licenses is not covered.

If you do today, IDA Pro Headless is the way to bring that usage into your agreement: keep your IDA Pro licenses for your analysts, and move your scripts and agents to headless seats. Your scripts only need `IDA_LICENSE` set to check out a seat. [Contact us](https://hex-rays.com/contact) to scope the number of seats you need.

## What's next?

Check the [Getting Started](https://docs.hex-rays.com/editions-and-licenses/license-types/ida-pro-headless/getting-started.md) guide to install the package and analyze your first binary.


---

# Agent Instructions
This documentation is published with GitBook. GitBook is the documentation platform designed so that both humans and AI agents can read, navigate, and reason over technical content effectively. Learn more at gitbook.com.

## Querying This Documentation
If you need additional information that is not directly available in this page, you can query the documentation by asking a question.

Perform an HTTP GET request on the following URL with the `ask` and `goal` query parameters:

```
GET https://docs.hex-rays.com/editions-and-licenses/license-types/ida-pro-headless/overview.md?ask=<question>&goal=<user_goal>
```

`ask` is the immediate question: it should be specific, self-contained, and written in natural language.
`goal` is what the user is ultimately trying to achieve, the reason they need the answer. Sharing it helps GitBook give you a better, more relevant answer. A goal is most helpful when it describes the outcome the user wants rather than restating the question. For example, with `ask=how do I create an API token`, a goal like `automate deployments from our CI pipeline` lets GitBook tailor the answer to that use case.

The response will contain a direct answer to the question and relevant excerpts and sources from the documentation.

Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.
