What is Jev Search for Obsidian?

An Obsidian plugin that searches your vault locally and offline with no API key, then optionally reranks the top results with Jev only after you approve exactly what gets sent. It is an experimental v0.1.0 preview.

How does the two-stage search work?

Stage one runs an on-device BM25 engine with a weighted hybrid tokenizer over the vault and extracts the top 20 candidates in about 15 ms; stage two lets Jev elevate the note that genuinely answers your question to number one.

What does Jev actually judge?

Jev evaluates whether a given note excerpt directly answers your query and returns calibrated probability distributions and structured scores. The README notes it does not generate conversational prose, so your notes stay intact.

How is privacy handled?

Candidates from excluded folders and #private tags are dropped before the consent step, a modal shows the exact JSON payload and cost before anything is sent, and the README cites TypeSafe/OpenRouter Zero Data Retention agreements.

How much does a rerank cost?

The README quotes $0.042 per million input tokens with output tokens free, and a measured 3-document rerank of about $0.000028, roughly $0.03 per 1,000 searches, based on the project's own live measurements.

How much does it improve results?

Its README benchmark of 120 live queries reports nDCG@10 rising from 0.5258 to 0.8788, a 67% improvement, with the cloud API round-trip taking about 750 milliseconds.

Do I need an API key to search at all?

No: local BM25 search works offline with no key. A key is only needed for the optional Jev rerank, and the digest does not include setup instructions because the README is truncated before that point.

Jev directory

CommunityJev Tools & CLIs

Jev Search for Obsidian

Obsidian plugin that searches a vault locally and offline with BM25 and no API key, then reranks the top candidates with Jev only after you approve exactly what gets sent. The README reports nDCG@10 rising from 0.53 to 0.88 on 120 queries.

See highlights Open github.com

Stars: 1
Language: JavaScript
License: MIT
Reading time: 2 min
Published by: Community
Author: jh1373
Last verified: Oct 2026

jh1373/jev-search

User repository on GitHub

View on GitHub

Search your Obsidian vault locally and offline with no API key, then rerank the top results with Jev only after you approve exactly what gets sent. Experimental preview.

Forks: 1
Last pushed: Updated Sep 2026

Repo stats from the GitHub API, cached Oct 2026.

Tagscommunity catalog-import madewithjev obsidian search

Works forJev for Ranking Jev for Classification

Used inJev for Software & DevTools

Languages: JavaScript
Capabilities: Open sourceRuns locallyRequires API key

Highlights

Two-stage search: a local weighted hybrid tokenizer with BM25 narrows 10,000 notes to the top 20 in about 15 ms, entirely offline.
Jev reranks only after a consent modal where you inspect the exact JSON payload and cost; excluded folders and #private tags never leave the machine.
README benchmark: nDCG@10 improved from 0.5258 to 0.8788 across 120 live queries, with cloud round-trips around 750 ms.
README pricing: $0.042 per million input tokens with free output, and a measured 3-document rerank costing about $0.000028, roughly $0.03 per 1,000 searches.
Ships English and Japanese documentation and cites TypeSafe/OpenRouter Zero Data Retention so notes are not stored or used to train models.

Watch out

Grounded on the truncated README and GitHub metadata reviewed 2026-10-15; the architecture section is cut off inside the privacy subgraph and the digest does not show installation steps or the consent modal itself. Benchmark and pricing figures are the project's own live measurements, not independent tests.

Questions about this entry

Short answers drawn from the linked project or page.

Reactions & coverage

Posts, threads, and videos about this entry from around the web.

X: LinkedIn bookmarks search

X: 1,066 reaction videos, searchable

X: Intent-based search in a Chrome extension

X: Margin, bookmark search

From the Jev guides

Original write-ups on what this entry is for.

Jev at home: voice pipelines, Assist routing, and sentinels

Oct 16, 2026

How Home Assistant conversation agents, macOS voice pipelines, and an advisory sentinel put a fast typed decision in the loop — one call per utterance, thresholds in code, and a daily token budget.

Jev Tools & CLIs#home-assistant#voice#automation

Read article

Search visibility, audited by decisions

Oct 16, 2026

Jev-based SEO tools split the crawl from the call: deterministic crawlers count and parse, while typed judgments rank fixes, place internal links, score search results, and track AI-answer mentions — with costs and caveats attached.

Jev Tools & CLIs#seo#search-visibility#audits

Read article

Jev on the Gateways: Vercel, Workers AI, and OpenRouter

Oct 12, 2026

The request shape barely changes across surfaces; the surrounding contract does. When to call the API directly and when a gateway, a Worker binding, or a router entry earns its extra hop.

Jev Tools & CLIs#gateway#vercel#cloudflare

Read article

From the Jev AI model community

Posts from builders shipping with Jev right now.

Follow @typesafeai

Jev Search for Obsidian

Highlights

Questions about this entry

What is Jev Search for Obsidian?

How does the two-stage search work?

What does Jev actually judge?

How is privacy handled?

How much does a rerank cost?

How much does it improve results?

Do I need an API key to search at all?

Reactions & coverage

Related terms

More Jev resources like this

From the Jev guides

From the Jev AI model community