> ## Documentation Index
> Fetch the complete documentation index at: https://arka-agent-mintlify-f21cea86.mintlify.site/llms.txt
> Use this file to discover all available pages before exploring further.

# Local LLM research notes

> Upstream local-model projects reviewed for Arka improvements.

Arka keeps shallow upstream checkouts in `.third_party/local-llm-research/` for
offline reference. Arka does not import them into the runtime, and
Git ignores them. The current research set is:

| Project | License | Useful Arka ideas |
| - | - | - |
| [Ollama](https://github.com/ollama/ollama) | MIT | Model pull/list/run lifecycle, local API health checks |
| [llama.cpp](https://github.com/ggml-org/llama.cpp) | MIT | Portable CPU/Metal/CUDA inference and quantized-model sizing |
| [vLLM](https://github.com/vllm-project/vllm) | Apache-2.0 | OpenAI-compatible serving, GPU batching, throughput-oriented deployment |
| [Aider](https://github.com/Aider-AI/aider) | Apache-2.0 | Repository map, edit/test loop, local Ollama integration |

Arka’s local-model selector already uses installed Ollama models and hardware
headroom. The next safe integrations are adapter-level: improve model
discovery, expose llama.cpp/vLLM health probes, and reuse repository-map and
verification patterns. Do not copy source code or vendor dependencies without
reviewing each repository’s license and notices; transitive model/runtime
licenses may differ from the project license.

The checkouts are shallow clones for research. Refresh them
before relying on version-specific behavior.


## Related topics

- [Arka — AI terminal agent documentation](/index.md)
- [AI agent guide — use Arka over MCP](/guides/ai-agents.md)
- [Configuration, env vars, and API key setup](/reference/configuration.md)
- [Source-backed research across profession domains](/guides/professions.md)
- [Personalize your Arka experience](/guides/personalize.md)


This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.