CAPE Sandbox
capesandbox.com
- Free plan
- Open source
- Windows
- Linux
- Web
- API
- Self-hosted
What CAPE Sandbox is
CAPE Sandbox is an open-source tool for analyzing suspicious files in isolated environments while recording behavior and collecting forensic artifacts. It can capture files created or changed during execution, network traffic in PCAP format, screenshots, and memory dumps. CAPE also performs dynamic malware unpacking, classifies unpacked payloads with YARA, and extracts configurations through static and dynamic analysis.
A Django web interface supports file submission, report browsing, and analysis searches. A REST API and Python submission functions let teams automate file and URL analysis. Documented targets include Windows executables and DLLs, PDFs, Office files, URLs, HTML, scripts, ZIP files, and Java JARs. Deployment runs each analysis in a fresh virtual machine and requires host and guest machines; GNU/Linux, preferably Ubuntu LTS, is the recommended host, and Windows 10 or Windows 11 23H2 is the recommended guest. CAPE is free and self-hosted, distributed without warranty, and its documentation places responsibility for use on the user.
Who it suits
CAPE suits security teams and researchers who need to inspect suspicious files or URLs and collect behavioral artifacts. It is intended for users able to set up and maintain host and guest machines.
Good to know, and what to watch
- Captures behavioral activity and forensic artifacts
- Unpacks malware dynamically and classifies payloads with YARA
- Supports file and URL analysis automation
- Includes a Django interface for reports and searches
- Documents integrations with Box-js, LibreNMS, and Suricata
- Requires both host and guest machines
- Changing setup-script packages can break the installation
- Distributed without warranty; use is the user's responsibility
Plans
Every plan and priceCAPE Sandbox
Public demonstration instance
Open source
Getting started
- Open the CAPE Sandbox website and review the installation documentation
- Prepare a GNU/Linux host, preferably Ubuntu LTS
- Set up a Windows 10 or Windows 11 23H2 guest machine
- Follow the installation guide without changing packages installed by its setup script
- Submit files or URLs through the Django interface, REST API, or Python functions
Limits to check
CAPE requires host and guest machines for its isolated virtual-machine architecture. The documentation warns that changing packages installed by the setup script can break the KVM, libvirt, and CAPE installation.
The verdict
CAPE combines isolated analysis, artifact collection, unpacking, and automation in a self-hosted tool. Its machine requirements and setup cautions matter, and users are responsible for how they use it.
Questions people ask
Is CAPE Sandbox free?
Yes. CAPE Sandbox is free, open-source software intended for self-hosted deployment.
Which platforms does it support?
The listed platforms are API, Linux, self-hosted, web, and Windows. The recommended deployment uses GNU/Linux as host and Windows 10 or Windows 11 23H2 as guest.
Can it analyze URLs?
Yes. URL analysis is supported through the documented targets and automation interfaces.
Does it provide an API?
Yes. CAPE offers a REST API and Python submission functions for automating file and URL analysis.
What integrations are documented?
Documentation covers Box-js, LibreNMS, and Suricata, and describes modular integration with external services.
Is CAPE Sandbox covered by a warranty?
No. It is distributed without warranty, and the documentation places responsibility for use on the user.
What the maker publishes
- AI clients
- The CAPE MCP server connects CAPE instances with MCP-compliant clients, with examples for Claude Desktop, Gemini CLI, and Antigravity. source
- API authentication
- The API supports token authentication through the api-token-auth endpoint and Authorization headers. source
- API throttling
- The documented default API throttle is five requests per minute and can be changed in api.conf. source
- Architecture
- A Linux host manages isolated Windows guest virtual machines for each analysis. source
- Automation
- CAPE offers a REST API and Python submission functions for automating file and URL analysis. source
- Behavior monitoring
- Windows analysis uses API-hook instrumentation and behavioral signatures. source
- Company origin
- The project history identifies Kevin O’Reilly at Context Information Security as the original CAPE developer. source
- Configuration extraction
- Static and dynamic malware configuration extraction is built in, with optional RATDecoders, DC3-MWCP, MalDuck, and MaCo frameworks. source
- Debugger
- Its debugger can be programmed with YARA signatures for custom unpacking or configuration extractors, anti-sandbox countermeasures, and instruction traces. source
- Debugger
- The debugger supports YARA-triggered breakpoints, instruction traces, memory dumping, and control-flow changes to counter evasion. source
- Deployment
- The documented architecture runs each analysis in a fresh isolated virtual machine; GNU/Linux, preferably Ubuntu LTS, is the recommended host, with Windows 10 or Windows 11 23H2 as the recommended guest. source
- Distributed deployment
- Master and worker nodes can be distributed through a REST API on port 9003 with X-API-Token authentication. source
- Documentation depth
- The documentation covers installation, guests, networking, cloud deployment, API usage, customization, distributed mode, and troubleshooting, while warning that some material may be outdated. source
- Dynamic analysis
- It captures behavioral activity, files created or changed during execution, network traffic in PCAP format, screenshots, and memory dumps. source
- Dynamic unpacking
- CAPE automatically extracts payloads from techniques including process injection, hollowing, doppelgänging, and in-memory decompression. source
- File artifacts
- Analyses record files created, modified, or deleted during execution. source
- Founded
- 2016 source
- Hypervisor extensibility
- Custom Python machinery modules can add unsupported virtualization backends; documented options include QEMU/KVM, LXC, Xen, OpenVZ, and VMware. source
- Independent listing
- AlternativeTo describes CAPE as free and open source and reported no user comments or reviews on its listing. source
- Input types
- Documented analysis targets include Windows executables, DLLs, PDFs, Office files, URLs, HTML, scripts, ZIP files, and Java JARs. source
- Integrations
- The documentation covers integrations with Box-js, LibreNMS, and Suricata, and describes CAPE's modular design for integrating external services. source
- Licence
- The Cuckoo-derived code is distributed under GNU GPL version 3; bundled components may use other licences. source
- Limit: configuration extraction
- A SURICON paper reports that configuration extraction does not work every time and can produce inconsistent results across parser frameworks. source
- Limit: deployment complexity
- WorldMetrics says setup and orchestration require platform and infrastructure knowledge, with noisy results unless analysts filter them. source
Checked against the maker's pages on 15 Oct 2026.