Two versions of Kalea. Both are local, private, embodied, and voice-first. The difference is how much capability lives under the hood.
Both are local, private, embodied, and voice-first. The difference is memory: how many minds she can hold ready at once.
| Kalea lite | Kalea Pro | |
|---|---|---|
| Price | $594 | $835 |
| Best for | Single users, tutoring, companion | multi-model optimization, multi-user households, agents, and swarms |
| Memory | 8GB LPDDR5 | 16GB LPDDR5 |
| Storage | 64GB eMMC | 256GB eMMC |
| Finish | Blue | Black |
| Models kept warm | One at a time | Fast + deep brain together · both brains resident |
| Brain switching | A reload — about 45 seconds | Instant |
| Small Council (multi-model) | Available, slower | Prewarmed — fast |
| Household swarm (forthcoming) | Joins the graph as one sealed node | Built for multi-node, multi-user households |
Ship updates, firmware notes, and ideas from the edge.