Obsidian Stats logoObsidian Stats

Synapse AI

by Fredrix
5
4
3
2
1
Score: 50/100
Install Code

Description

Your second brain in Obsidian.

Reviews

No reviews yet.

Stats

2
stars
237
downloads
0
forks
95
days
95
days
95
days
0
total PRs
0
open PRs
0
closed PRs
0
merged PRs
0
total issues
0
open issues
0
closed issues
4
commits

Latest Version

0.5.2
3 months ago

Changelog

First public beta. Core chat with your notes works and is in daily use.

Features

  • 💬 Chat in a side panel — streaming responses, stop mid-stream, regenerate, edit your messages
  • 🔌 Bring your own model — Gemini, OpenRouter, or any OpenAI-compatible endpoint (custom baseURL, incl. local Ollama / LM Studio); add several and switch on the fly
  • 📎 Vault-aware context — pull notes in via [[wikilinks]], the active note, or automatic keyword search
  • 💾 Save reply as a note — turn any answer (or a selected fragment) into a permanent note
  • 🗂 Chat autosave — conversations saved as plain .md in your vault
  • 🌐 Proxy support for geo-blocked regions
  • 🌍 English & Russian UI

Install

  • Manual: download main.js, manifest.json, styles.css below → drop into <vault>/.obsidian/plugins/synapse-ai/ → enable in Community plugins
  • BRAT: add Fredrix1/synapse-ai-obsidian

⚠️ Beta — expect rough edges. Semantic search, agents and a local privacy layer are on the roadmap.

See all version on GitHub

README file from

Github

Your AI second brain inside Obsidian — chat with your notes, bring your own key, keep your data private.

Version Obsidian License

Report a bug · Request a feature · 🇬🇧 English · 🇷🇺 Русский

Synapse AI turns your vault into a knowledge base you can talk to. Ask questions, pull notes in as context, and save the best answers back as permanent notes — powered by the model you choose (Gemini, OpenRouter, or any OpenAI‑compatible API, including local Ollama / LM Studio) with your own key.

⚠️ Beta. Core chat works and is in daily use. Semantic search (embeddings), agents and a local privacy layer are on the roadmap.

Features

  • 💬 Chat in a side panel — streaming responses, stop mid‑stream, regenerate, edit your messages.
  • 🔌 Bring your own model — Gemini, OpenRouter, or any OpenAI‑compatible endpoint (custom baseURL, incl. local Ollama / LM Studio). Add several and switch on the fly.
  • 📎 Vault‑aware context — pull notes into the prompt via [[wikilinks]], the active note (toggle), or automatic keyword search.
  • 💾 Save reply as a note — turn any answer (or a selected fragment) into a permanent, linkable note with frontmatter. Right‑click a message for the full menu.
  • 🗂 Chat autosave — conversations saved as plain .md in your vault (configurable, optional).
  • 🌐 Proxy support — route requests through your own proxy/VPS (for geo‑blocked regions)
  • 🌍 Localization — English & Russian UI.
  • 🔒 You stay in control — your notes and key go only to the AI provider you pick (or your own proxy). No middle‑man server, no analytics.

Getting Started

Install — download manifest.json, main.js, styles.css from the latest Release into <vault>/.obsidian/plugins/synapse-ai/, reload Obsidian, enable Synapse AI in Community plugins. Or via BRAT: Add beta plugin → Fredrix1/synapse-ai-obsidian.

Setup

  1. Settings → Synapse AI → API Keys — add a key (OpenRouter / Gemini).
  2. Models → Add model — pick a provider (or a custom OpenAI‑compatible baseURL, e.g. local Ollama) and the model id; set your Default Chat Model.
  3. (Optional) Proxy URL for geo‑blocked regions; System Prompt, chat folder, default tag, language.

Open the chat via the command palette → “Synapse: Open Chat”. Type a question (Enter to send), add context with [[Note name]], and right‑click a reply → Save as note.

License

AGPL‑3.0 · by Fredrix · issues & ideas welcome on GitHub.

Sponsors