Both editions run the same inference engine and models through the same OpenAI-compatible API. Enterprise adds what an organization needs to run that across many PCs.
| Open source | Enterprise | |
|---|---|---|
| Only in Enterprise | ||
| Hosted console and certificate authority | — | ✓ |
| Self-enrolling devices, install links and batch installs | — | ✓ |
| API tokens with expiry and revocation | — | ✓ |
| Fleet Manager dashboard | — | ✓ |
| Signed automatic updates with rollback | — | ✓ |
| Relays for devices behind NAT | — | ✓ |
| Team members and roles | — | ✓ |
| Different in each | ||
| Model deployment across the fleet | Manual, per machine | From the Fleet Manager |
| Platforms | Windows and Linux | Windows 10/11 and Ubuntu 22.04+ |
| License | Apache-2.0 | Commercial, priced per device |
| Support | Community, via GitHub | Your Cascadia representative |
| In both editions | ||
| OpenAI-compatible API | ✓ | ✓ |
| Single- and multi-machine inference | ✓ | ✓ |
| Intel CPU, GPU and NPU | ✓ | ✓ |

