---
title: "llama.cpp server"
description: "llama.cpp server as a MainSDK driver for ai. Pick it in main.yaml; your calls stay the same."
section: "integrations"
canonical: "https://mainsdk.dev/docs/integrations/llama-cpp"
updated: "2026-10-28"
---

# llama.cpp server

> llama.cpp server as a MainSDK driver for ai. Pick it in main.yaml; your calls stay the same.

Canonical: https://mainsdk.dev/docs/integrations/llama-cpp · Updated: 2026-10-28

llama.cpp server as a MainSDK driver for ai. Pick it in main.yaml; your calls stay the same.

## Interfaces it serves

### AI calls (`ai`)

- Status: Building for 0.3 · ships as a beta API
- Driver: `llama-cpp` (llama.cpp server). not built yet
- Docs: https://mainsdk.dev/docs/interfaces/ai.md

```yaml
ai:
  driver: llama-cpp
```

Swap to another provider: `local`, `anthropic`, `openai`, `gateway`, `bedrock`, `ollama`, `openai-compatible`.