Ctrl K
ComunidadGuías…
GITHUB · ANALIZADO

obsidian-lilbee

lilbee es un plugin de Obsidian que funciona como motor de búsqueda y chat local para tu biblioteca de notas. Descarga y ejecuta modelos de IA locales en una sola clic, o usa Ollama/LM Studio. Responde preguntas con citas directas a las líneas fuente.

Ficha actualizada 22 sept 2026, 18:37 · upstream 22 sept 2026, 17:34 · release 0.6.97

Ver códigoProblemas y ayudaRepositorio ↗Sitio oficial
00

Qué es obsidian-lilbee

Descripción editorial construida a partir del repositorio, su documentación y sus señales de mantenimiento.

Obsidian community plugin: your vault as a private, self-hosted NotebookLM alternative. A built-in manager downloads and runs local AI models in one click, or use your Ollama and LM Studio instances. Cites answers from notes, PDFs, and code repositories. Crawls websites into your library with full offline support. Your own models live on your machine with persistent memory that remembers how you like to be answered.

CASOS DE USO
  • Ask questions about your vault contents (notes, PDFs, code) and get cited answers
  • Crawl external websites and save them as searchable markdown notes in your vault
  • Build a personal encyclopedia from your own documents without needing paid APIs
  • Run AI models locally on any GPU to answer complex questions with full transparency
  • Create an auto-generated wiki from indexed content that stays synchronized with your vault
PUNTOS DESTACADOS
  • All local inference runs on your machine — no cloud costs, no API keys needed for basic use
  • Built-in web crawler saves any website to your vault as searchable markdown notes
  • GPU placement view lets you assign roles (chat, embed, rerank) across multiple GPUs
  • Memory feature remembers your preferences and recalls relevant context in future chats
  • Works with Ollama or LM Studio — both model managers appear alongside lilbee's own catalog
  • Every answer includes inline citations that open directly to the source line
PRECAUCIONES
  • Cloud models are opt-in per role and require an active subscription for those features
  • Model size can be significant (hundreds of MB to several GB) depending on what you choose
  • Offline crawling requires internet access during the initial crawl phase, though saved notes remain local
  • No built-in document-level citation — citations are per-line and must be manually verified
  • Memory is opt-in and managed through a separate Memories view rather than being automatic
01

Perfil de infraestructura

Estimación inicial derivada del repositorio y de los servicios detectados.

CPU4 vCPUrecomendación inicial
MEMORIA8 GBRAM estimada
DISCO10 GBalmacenamiento base
COMPLEJIDADlownivel de operación
02

Despliegue detectado

Qué señales reales de instalación hemos encontrado.

Despliegue manual

Install via Obsidian Community Plugins. No server setup needed for basic use. For multi-GPU setups, configure GPU placement in the model selection UI. If using Ollama or LM Studio, ensure they are running before launching lilbee. The plugin handles all model downloads and serving internally — no additional software installation required.

—Dockerfile
—Compose
—One-click
03

Stack y dependencias

Componentes inferidos a partir de manifests, configuración y documentación.

FRAMEWORKS
Next.jsNode.js
SERVICIOS
No detectados
BASES DE DATOS
No detectadas
PUERTOS
7433
05

Puede sustituir a

Relaciones detectadas por el análisis del repositorio y su categoría.

06

Experiencia real de la comunidad

Ayuda a contrastar la estimación automática con instalaciones reales.

Todavía no hay datos comunitarios para este proyecto.