Obsidian Copilot
The Knowledge Work Agent Designed for Your Second Brain
Copilot is the AI plugin to beat inside Obsidian, mostly because it refuses to trap you: the free tier runs on your own keys or existing ChatGPT/Claude subscription, and meaning-based vault search is the part that actually changes how you work. V4's Agent Chat — opencode, Claude Code, or Codex reading your vault with approval gates — is a real step past chat-in-a-sidebar. Where it earns money is the managed layer: hosted models, multi-agent answers, OpenArtifacts publishing, Miyo, and self-hosting, which starts at the $349.99 one-time Supporter tier. If you don't live in Obsidian, none of this matters.
Verified 4d ago · liveness 73/100 · cite: rightaichoice.com/tools/obsidian-copilot
- Obsidian power users who want AI woven into note-taking, not a separate chat app
- Researchers who need meaning-based search across a vault of notes, PDFs, and EPUBs
- Writers who want rewrites, expansions, and cited answers without leaving Markdown
- Privacy-conscious users who want local models and self-host mode (Supporter)
- Anyone who doesn't use Obsidian — the plugin has no standalone value
- Teams wanting a shared, browser-based knowledge base rather than a local vault
- Mobile-first users who need Agent Chat, which is unavailable on mobile
We scan live Reddit threads, YouTube comments, X posts, G2 reviews and other communities — and hand you an honest verdict in under a minute.
- Honest verdict, not marketing
- Real pros & cons from real users
- Attributed quotes with receipts
3 free scans · no card needed
Skip Copilot for Obsidian if you don't use Obsidian at all, or if you need Agent Chat on mobile — it's desktop-only and the plugin has no standalone value outside the vault.
Copilot-hosted models on Lite and Plus are subject to 5-hour and weekly usage limits, so a heavy research week can hit the cap before the month is out.
Free (BYOK) suits anyone already paying for ChatGPT or Claude. Lite at $6.25/mo billed yearly ($74.99) is the entry to hosted models but excludes Plus-only features like multi-agent answers and OpenArtifacts. Plus at $11.67/mo billed yearly ($139.99, versus $14.99 monthly) is the tier most buyers want. Supporter at $349.99 one-time bundles 2 years of Plus plus lifetime Miyo ($149.99) and OpenArtifacts ($69.99).
In short
Obsidian Copilot — The Knowledge Work Agent Designed for Your Second Brain. Best for Obsidian power users who want AI woven into note-taking, not a separate chat app, Researchers who need meaning-based search across a vault of notes, PDFs, and EPUBs, Writers who want rewrites, expansions, and cited answers without leaving Markdown. Free to start; paid plans from $6.25/mo.
What people actually say about Obsidian Copilot — is it worth it?
We ran a structured research pass across product reviews, community discussions, and post-purchase forum threads to surface the patterns vendors won't publish themselves. Below: the recurring strengths, the hidden costs people mention most, and the cohort that consistently regrets adopting this tool.
7 mentions across 3 sources (Hacker News, GitHub, Lemmy) · researched Jul 3, 2026.
Average across the 3 sources that answered — each source counts once, not each post.
- +Deep Obsidian integration—everything stored as Markdown files with no lock-in.
- +Supports any OpenAI-compatible LLM, including local via LM Studio/Ollama.
- +Vault QA with cited answers from your notes and live sources.
- +Agentic AI with web search, long-term memory (Plus).
- +Composer for rewriting notes with diff review before accepting changes.
- −Free tier lacks core features like AI agents and PDF support.
- −Requires your own API key—costs can add up with cloud models.
- −No native knowledge graph visualization.
- −233 open GitHub issues suggest support or reliability gaps.
- −Tightly coupled to Obsidian—zero value outside that ecosystem.
- • Free tier requires OpenAI/other API key—costs vary by usage
- • Self-Host Supporter is a one-time high fee ($349.99)
Viability Score
How well maintained and how widely used is Obsidian Copilot? Built from what the vendor actually publishes (docs, changelog, tutorials, integrations, pricing), whether the site is live, and how much real users discuss it. How we calculate this
Last calculated: October 2026
How we score →Key Features
- Chat with any LLM: OpenAI, Anthropic, Google, LM Studio, Ollama, or OpenAI-compatible endpoints
- Agent Chat with opencode, Claude Code, or Codex reading your vault with approval gates
- Multi-agent answers: several agents research the same question and return one summary
- Vault QA with cited answers from your notes and live web
- Projects: isolated workspaces with dedicated model, instructions, and chat history
- Semantic vault search: find notes by meaning, not keywords
- Local indexing via Miyo, embedding notes plus outside PDFs and EPUBs
- Live Relevant Notes surface automatically, ranked by meaning
- Speaks Obsidian natively: wikilinks, canvases, plain Markdown output
- Copilot Commands and Quick Ask at cursor, right-click menu, or command palette
- Quick Chat for lightweight conversations and the main mobile chat view
- Context and Mentions: add notes, selections, folders, URLs, and other agents to a request
- Skills shared across agents: opencode, Claude Code, and Codex
- AGENTS.md and instruction files for durable per-vault and per-project guidance
- Self-host mode: run local models for full privacy (Supporter)
About Obsidian Copilot
Copilot for Obsidian is a plugin that turns your vault into an AI knowledge work agent. The free plugin runs on your existing ChatGPT or Claude subscription, or your own API keys — OpenAI, Anthropic, Google, LM Studio, Ollama, or any OpenAI-compatible endpoint — so you're never locked into one provider. Search is the headline act: ask for a note the way you remember it, half an idea in plain words, and the agent searches your vault by meaning, returning the exact match plus related captures you'd forgotten. Related notes surface automatically beside whatever you're writing, powered by Miyo, a local indexing engine that embeds your notes along with outside PDFs and EPUBs. Version 4 is built around Agent Chat, where opencode, Claude Code, or Codex read your vault, use tools, and complete multi-step work with your approval; Quick Ask, Copilot Commands, and Quick Chat remain for smaller jobs and mobile use. Output lands as Obsidian-native wikilinks, canvases, and plain Markdown, so work becomes notes you keep rather than chat transcripts that scroll away. Skills are reusable — add one once and flip it on per agent across opencode, Claude Code, and Codex. Paid licenses add the managed layer: Copilot-hosted models with no API key, multi-agent collaboration, better web search and PDF parsing, built-in YouTube and X capture, and OpenArtifacts one-click publishing.
Behind the Verdict
Strengths. The provider story is unusually honest: the free plugin asks for no sign-up and no license key, and runs on the ChatGPT or Claude subscription you already pay for or on your own API keys — including local endpoints through LM Studio and Ollama. Search is the feature that justifies the install. Ask 'where's that note about attention only reading some of the tokens?' and Copilot returns [[Sparse attention]] with the quote, then points at the two other notes on the same idea. That is meaning-based retrieval over your vault plus outside PDFs and EPUBs, and it runs on Miyo, a local indexing engine. Miyo is also sold standalone and can be connected to Copilot, including from several devices via a shared host with Tailscale. V4 reframes the plugin around Agent Chat. opencode, Claude Code, or Codex can read the vault, use tools, and complete multi-step work with your approval; Copilot installs and manages opencode and the Codex adapter, or auto-detects Claude Code. Most people should start with opencode. Projects give ongoing work their own instructions, context, and chat history. Context and Mentions let you pull notes, selections, folders, URLs, and other agents into a request, and AGENTS.md files carry durable instructions. Skills are written once and reused across all three agents, which is the detail most multi-agent setups get wrong. Weaknesses and honest trade-offs. This is a plugin, not an app — without Obsidian it has no standalone value. Agent Chat is unavailable on mobile, where Quick Chat is the main view. Self-host mode requires Supporter; Lite and Plus explicitly have no self-hosting support, which is an odd shape for privacy-first buyers, though Supporters keep lifetime self-host mode even after two years. Hosted-model usage is metered: Lite is subject to 5-hour and weekly quotas, Plus includes 3x Lite's allowance, and past the cap you buy Copilot credits. Copilot publishes its own hosted-model benchmark chart, and the footnote is worth reading — 'Copilot Plus Flash' is a 56.2 internal estimate because Artificial Analysis does not list that hosted alias. Where it fits. Researchers, writers, engineers, and analysts whose second brain already lives in Markdown. Where it doesn't: teams wanting a shared browser-based knowledge base, or anyone expecting a fully free assistant without leaning on an existing AI subscription or API keys.
Researching Obsidian Copilot? Get your full AI stack in 60 seconds.
Free, no signup — tell us your goal and get tools matched to your budget & existing stack.
Real-world workflow fit
Concrete scenarios for the personas Obsidian Copilot actually fits — and what changes day-one when you adopt it.
You remember an idea but not the filename, so you describe it in plain words in Agent Chat. Copilot searches the vault by meaning, returns the source note, and flags two earlier captures of the same idea.
Outcome: You find the note and rediscover two related captures you'd forgotten, with wikilinks you can drop straight into the current draft.
Quick Ask at the cursor pulls one note into context, Live Relevant Notes surfaces a related note beside the draft, and Composer rewrites a paragraph for review before you apply it.
Outcome: The rewrite lands as Markdown in the note you were already writing, with the related note linked rather than lost.
You open Agent Chat with opencode, point it at a week of captured readings, and approve each step as it folds them into evergreen wiki notes and flags a claim newer sources contradict.
Outcome: The reading queue becomes linked wiki notes in one session, with the contradiction surfaced instead of buried.
Use Cases
- Ask questions about your notes and get cited answers from your vault.
- Find a half-remembered note by describing it in plain words and get the exact match plus related captures.
- Run multi-step vault work in Agent Chat with opencode, Claude Code, or Codex, approving each step.
- Put several agents on the same research question and get one unified summary.
- Organize ongoing work into Projects with their own instructions, model, and chat history.
- Run Obsidian commands and reusable prompts via chat or inline Quick Ask.
- Pull YouTube videos, X threads, and web pages directly into your chat context.
- Publish a Markdown note at one stable public link with version history via OpenArtifacts.
Models Under the Hood
as of 2026-09-22
Limitations
- Requires Obsidian as the host environment — it's a plugin, not a standalone app.
- Agent Chat is unavailable on mobile; Quick Chat is the main chat view there.
- The free tier is BYOK, relying on your existing ChatGPT/Claude subscription or your own API keys, with no hosted models or license key required.
- Paid Lite and Plus plans add Copilot-hosted models but are subject to 5-hour and weekly included usage limits, and to token allowances (Plus includes 3x Lite's allowance); past the cap you buy Copilot credits.
- Multi-agent collaboration, better web search and PDF parsing, YouTube/X capture skills, and OpenArtifacts publishing require Plus; Plus has no self-hosting support.
- Self-host mode requires Supporter, though Supporters keep lifetime self-host mode after the included 2 years of Plus end.
- Windows Agent Chat setup has its own documented path.
as of 2026-10-03
Verification history
We have re-verified Obsidian Copilot 8 times since . Each pass re-reads the vendor's own pages and re-checks every listed field against that evidence; passes where nothing had changed are marked as such.
- — re-verified summary, description, our verdict, our analysis, pricing model, pricing tiers, features, integrations, who it suits, who should skip it
- — re-verified summary, description, our verdict, our analysis, pricing model, pricing tiers, features, integrations, who it suits, who should skip it
- — re-verified summary, description, our verdict, our analysis, pricing model, pricing tiers, features, integrations, who it suits, who should skip it
- — re-verified summary, description, our verdict, our analysis, pricing model, pricing tiers, features, integrations, who it suits, who should skip it
- — re-checked, vendor evidence unchanged
- — re-verified summary, description, our verdict, our analysis, pricing model, pricing tiers, features, integrations, who it suits, who should skip it
Showing the 6 most recent of 8 verification passes.
Free to cite with attribution — this page re-verifies continuously.
12-month cost
Project the real annual outlay, including the implied monthly cost when only an annual tier is published.
Vendor list price only. Add-on usage, seat overages, and contract minimums are surfaced under Hidden costs & gotchas.
Plans compared
For each published Obsidian Copilot tier: who it actually fits, and what it adds vs. the previous tier. Cross-reference the cost calculator above for projected annual outlay.
Free
$0
Ideal for
Obsidian users already paying for ChatGPT or Claude, or anyone with their own API keys or a local model
What this tier adds
Free entry point — BYOK with no sign-up or license key, no hosted models
Lite
$6.25/mo (billed yearly at $74.99)
Ideal for
Solo users who want Copilot-hosted models without setting up a provider account or API key
What this tier adds
Adds Copilot-hosted models, subject to 5-hour and weekly usage limits; multi-agent collaboration, better web search and PDF parsing, YouTube/X skills, and OpenArtifacts still require Plus
Plus
$11.67/mo billed yearly ($139.99/year; $14.99/mo monthly)
Ideal for
Researchers and writers who want multi-agent answers, better web search and PDF parsing, and OpenArtifacts publishing
What this tier adds
Adds 3x Lite's token allowance plus multi-agent collaboration, better web search and PDF parsing, YouTube and X capture skills, and OpenArtifacts publishing — with no self-hosting support
Supporter
$349.99 one-time
Ideal for
Privacy-first users who want lifetime self-host mode and long-term ownership of Miyo and OpenArtifacts
What this tier adds
Adds 2 years of Copilot Plus, lifetime self-host mode, early feature access, 15% extra credit on usage purchases, plus lifetime Miyo and OpenArtifacts
Where the pricing makes sense
The company stage and team size where Obsidian Copilot's pricing actually pencils out — and where peers do it cheaper.
Free (BYOK) suits anyone already paying for ChatGPT or Claude. Lite at $6.25/mo billed yearly ($74.99) is the entry to hosted models but excludes Plus-only features like multi-agent answers and OpenArtifacts. Plus at $11.67/mo billed yearly ($139.99, versus $14.99 monthly) is the tier most buyers want. Supporter at $349.99 one-time bundles 2 years of Plus plus lifetime Miyo ($149.99) and OpenArtifacts ($69.99).
Setup time & first value
How long it actually takes to get something useful out of Obsidian Copilot — broken out by persona, not the marketing-page minute.
Free tier: install from Obsidian's Community Plugins, add your API key or connect an existing ChatGPT/Claude subscription — usually under 15 minutes. Agent Chat: open Settings → Copilot → Basic → Agents, download managed opencode or Codex (or auto-detect Claude Code), then run Open Copilot Agent Chat Window; most people should start with opencode. Miyo setup is walked through with screenshots and
Switching to or from Obsidian Copilot
How to bring data in from common predecessors and how to get it back out — written for the switcher, not the buyer.
- →From Obsidian's built-in search: point Miyo at the vault once and start querying by meaning instead of keywords — existing folders and tags stay untouched.
- →From a separate chat app: install the plugin, open Agent Chat, and your notes become the context source instead of pasted excerpts.
- →From Claude Code or Codex alone: auto-detect or install the adapter so both agents read the vault, and move shared instructions into Skills and AGENTS.md.
- →From a standalone Miyo install: connect the existing Miyo server to Copilot, optionally sharing one server across devices via Tailscale.
- ↗To a browser-based team wiki: vault notes export as plain Markdown, but you lose semantic search, Agent Chat, and Live Relevant Notes in the move.
- ↗To a standalone chat assistant: export notes and paste context manually — you trade cited vault answers for a polished chat UI.
- ↗To Miyo standalone: Copilot's indexing engine is sold separately, so local-first search survives even if you drop the plugin's agent layer.
Integrations
Resources & Guides
- Documentationobsidiancopilot.com
Docs · Obsidian Copilot
Full product docs from obsidiancopilot.com
- Quickstartobsidiancopilot.com
Getting Started · Obsidian Copilot
Get up and running fast from obsidiancopilot.com
- Documentationobsidiancopilot.com
Settings · Obsidian Copilot
Full product docs from obsidiancopilot.com
- Resourceobsidiancopilot.com
Changelog · Obsidian Copilot
Helpful link from obsidiancopilot.com
Tutorials & Learning
The Obsidian Copilot Plugin: Truly Automate Your Knowledge Vault (Save Time)
Veni, vidi, vici.
Obsidian 🤖 Copilot プラグイン + LM Studio 📝 でローカル AI を実行する方法
Wanderloots
4 Ways to Chat with Your Notes - Obsidian Copilot
Logan Hallucinates
YouTube returned 6 videos for “Obsidian Copilot”, and we withheld 2: 2 did not mention Obsidian Copilot. Showing the 4 we can prove are about Obsidian Copilot.
Official links
Tools that pair well with Obsidian Copilot
Common stack mates teams adopt alongside Obsidian Copilot, with the specific reason each pairing earns its keep.
ChatGPT
ChatGPT is OpenAI's AI chat assistant for reasoning, Codex coding, deep research, voice and agentic work on one subscription.
ChatPDF
ChatPDF turns any PDF into a chat partner: ask questions, get cited answers, summaries, and translations in seconds.
Heptabase
A visual knowledge base with an AI Tutor that builds structured courses from your own sources.
Featured Head-to-Head Comparisons
Obsidian Copilot vs Praktika
Choose Praktika if you're an intermediate language learner wanting immersive speaking practice with instant feedback. Choose Obsidian Copilot if you're an Obsidian user wanting AI-powered vault search, content generation, and automation. They serve entirely different use cases—language fluency vs productivity—so the choice depends on your primary need.
Obsidian Copilot vs Surge Ai
Obsidian Copilot and Surge AI serve completely different needs. If you're an Obsidian user seeking a deep AI integration for personal knowledge management, Copilot's free plugin with local model support is unbeatable. If you're a frontier AI lab requiring expert human feedback for RLHF or red teaming, Surge's curated workforce and recent Microsoft partnership prove its value. Choose based on whether you need to augment your own notes or train smarter models.
Alternatives to Obsidian Copilot
View all
Frequently Asked Questions
Best-of guides
Used Obsidian Copilot? Help shape our editorial sentiment research.