Show HN: Selora – local model for Home Assistant

Show HN: Selora – local model for Home Assistant(github.com)

7 points by bayshark 30 days ago | 4 comments

Selora AI Local is an open-source, Qwen-based model for Home Assistant.

Specs: Qwen3 1.7B base model (Q6 quantized~1.6GB) Four Home Assistant-specific LoRA adapters: - Answers - Clarifications - Automations - Commands ~3.5 GB total download size Runs locally via llama.cpp

We chose a Qwen-based architecture because of a paper on Arxiv (link below) which applied a Qwen based model for local LLM configuration, and showed promising results. We took it a step further in application by training LoRA adapters specialized in Home Assistant configuration.

This alpha release ships our base model with four specialist LoRA adapters preloaded for faster response: answers, clarifications, automations, and commands. The Q6 quantized base model is 1.6GB and the adapters are less than 100MB running either on self-hosted llama.cpp, or on Selora Hub devices, where everything is preconfigured and plug-and-play for you.

We started working on this because the existing options for local LLMs in Home Assistant lack knowledge to run anything useful, so users opt for very large LLMs, typically a cloud model that’s too expensive and not optimized for the kind of high-frequency, always-on smart home use we care about. We think there's a need for open-source models that are small and specialized enough to run on the kind of hardware people actually have at home.

Would love any feedback, questions, or ideas. Thanks for checking it out!

Hugging Face: https://huggingface.co/selorahomes/Selora-AI Arxiv: https://arxiv.org/abs/2502.12923 More details: https://selorahomes.com/selora-ai/

Anthonious 30 days ago |

Would be interesting to see how a specialised model does against a model with more params, I use qwen8b in my HA and would love to downgrade to 1.7B and optimise HA... But you have no stats or comparisons

bayshark 30 days ago | |

Definitely! We’re working on a benchmark in the coming weeks, and although we don't have data to support against qwen8b if you do try it out please let us know your experience in how it compares.

evil-olive 30 days ago |

it's a "local model" but your pricing page [0] has "Selora AI credits/month" and I get more credits if I pay more per month.

are you trying to get me to pay for "AI credits" even though the model is running on hardware I own?

0: https://selorahomes.com/pricing/

bayshark 30 days ago | |

We have two models: a cloud version that we released ~1.5 mos. ago and now the local model which took us a bit more time to release. The credits only apply to the cloud version. With our cloud version, you start off with 5,000 free credits and can purchase add-on credits. You can also bring your own API keys if you want to use a commercial cloud model. For local, none of that applies. Thanks for the comment, and we need to update that page to avoid confusion.