Get Pro $49
Start free with local chat and custom guides. Add natural spoken replies with Pro. Use Ultimate R7.7.1 for Guide conversations, saved replies and local voice with its bundled primary engine or optional LM Studio backend.
Find the features that fit the way you work.
Three editions, built around the features you use. Compare them below and choose your starting point.
Typed local chat through LM Studio with configurable Spirit Guides and local conversation history. You supply the model.
Your first conversation.Start exploring local AI, with a Guide you make your own.
SpiritBox concept artworkType to your local model. Hear its answer in a natural Piper voice you choose. Custom guides, local history, and a two-service Docker setup with LM Studio.
A voice that feels like yours.Shape the perspective. Choose the voice. Keep the conversation going.
SpiritBox concept artworkGuide-based AI conversations, saved conversations and local spoken replies. R7.7.1 setup guidance starts with Minimal and the bundled spiritbox-llm engine. LM Studio is optional. Broader routes remain unqualified.
More of your world, connected.Bring knowledge, tools, and creative workflows into your local workspace.
SpiritBox concept artworkBasic 1.3.2 downloads are available for Windows and macOS Apple Silicon. Both require LM Studio, a compatible local model, and its server on port 1234; no Docker is required. The compiled Windows build embeds Python and application libraries, so it does not require a separate Python installation. For Mac setup, follow the instructions included with the DMG. Intel Mac support is not offered.
SpiritBox Pro runs as two Docker services: the conversation app and Piper. On Windows 10/11 x64, use hardware virtualization and Docker Desktop with Linux containers and WSL2. On macOS, use an Apple Silicon Mac and Docker Desktop for Apple Silicon; WSL2 is Windows-only. Both downloads require LM Studio with a compatible local model and its server on port 1234. Model files are not included. First installation may download image and package dependencies.
Basic and Pro do not include microphone input, document memory, voice cloning or voice training.
Ultimate R7.7.1 is Windows Early Access for technical users with Docker Desktop using Linux containers and WSL2. The bundled spiritbox-llm engine is primary; LM Studio is optional. Models are separate. Start with Minimal and QUICK_START.md; LMSTUDIO_FIRST.md covers the optional backend. Prior-release bounded bundled-engine and optional LM Studio chat checks passed on a configured host, including saved conversations and restart persistence. Full and other hardware configurations remain unqualified. ComfyUI and n8n are not supplied as a configured, tested workflow bundle.
Check setup requirementsFind the right mix of conversation, voice, knowledge, and creative tools. Ultimate 1.12 R7.7.1 · Early access.
| Capability | Basic | Pro | Ultimate |
|---|---|---|---|
| Local model chat | LM Studio | LM Studio | Bundled engine primary; LM Studio optional |
| Spirit Guides | Core guides | Custom Guide Studio | Guides + saved conversations |
| Voice output | Browser/system voice | Piper TTS | Local Piper; Kathleen included |
| Voice training | Not included | Not included; compatible Piper voices for playback | Optional; unqualified in R7.7.1 |
| Stored context | Local conversation history | Local conversation history | Saved conversations; Full knowledge route unqualified |
| Document recall | Not included | Not included | Full route unqualified |
| ComfyUI | Not included | Not included | Optional, user-configured; unqualified in R7.7.1 |
| Automation | Not included | Not included | User-managed n8n; no configured workflow bundle |
| System complexity | Windows / macOS Apple Silicon + LM Studio | 2-service Docker stack + LM Studio | Minimal profile + bundled engine or optional LM Studio |
| Regular price | $0+ USD | $49 USD one-time | $199 USD one-time |
R7.7.1 hardware needs depend on your model, context length and configuration with the bundled engine or optional LM Studio. No general RAM or VRAM minimum is established by this release; other hardware and broader workstation routes remain unqualified.
View full requirements