Tech USP: All AI calls go through one OpenAI-compatible adapter. Anthropic, OpenAI, Llama via Ollama, vLLM, on-prem setups — interchangeable via environment variables, no code change.
Three modes
| Mode | When | Setup |
|---|---|---|
| Cloud API | kumiko.so hosted default | Anthropic API with platform key |
| BYOK Cloud | Power users with their own Anthropic / OpenAI account | Env vars (LLM_ENDPOINT / LLM_API_KEY / LLM_MODEL) |
| Local model | European mid-market on-prem | Ollama / vLLM on your own GPU server |
Killer argument for European customers
“Your data and your AI never leave your data center.”
No other AI builder on the market offers this. Lovable, v0, Bolt, Retool-AI are all US-cloud only.
Architecture deep dive
- Architecture decision: OpenAI-compatible adapter — single integration point for every provider
- LLM gateway, BYOK, local-LLM docs in the platform repo
Where this lands in the pitch
- EU mid-market: Top argument “data and AI stay with you” — own server, own model
- Indie hackers: Sub-argument “switch providers anytime, no lock-in”