Vellum

Your private, memory-first AI assistant that lives on your machine.

Intermediate
Visit VellumBack to AI tools

Own Vellum? Verify ownership to get a dofollow link

Screenshot of Vellum, Your private, memory-first AI assistant that lives on your machine

What is Vellum?

Vellum is a private, memory-first AI assistant that runs locally or in the cloud, connects to Google Workspace, Slack and Notion, and takes actions across your tools with granular permissions. It supports multiple model providers and is reachable from Mac, web, mobile, voice and chat apps.

Vellum is a personal AI assistant built around long-term memory and local ownership. Unlike general chat assistants, it is designed to remember your life, people, preferences and work across every conversation, and to act on that context without being prompted each time. The assistant runs locally on your machine by default, with conversation history, memory and credentials stored on your device and never used to train models. Vellum is open-source, so the code can be inspected, and cloud hosting runs in an isolated container provisioned for a single user. The assistant is model-agnostic. You can switch between OpenAI, Anthropic, Gemini or open-weight models via Ollama at any time, which makes it a practical hub for people who already experiment with different providers. Built-in skills cover Google Workspace, Slack, Notion and the local filesystem, and the system can be extended with community plugins or custom skills. Typical work includes drafting emails, managing calendars, building slides, writing code, running research and monitoring websites on a recurring schedule. Vellum is aimed at knowledge workers, founders and operators who want an assistant that behaves like a long-term colleague rather than a stateless chatbot. It is native to Mac and also reachable through iOS, Android, the web app, voice, email, Telegram, Slack and the terminal, with the same memory and task state across surfaces. A fine-grained permissions layer gates high-stakes actions (emailing new contacts, editing important files, spending credits) until you mark them as trusted. Among alternatives, Vellum sits closest to autonomous agent platforms and privacy-first personal assistants rather than to ChatGPT or Claude. It competes on memory persistence, local data control and the ability to chain actions across tools, and it fits users who are comfortable configuring permissions and integrations. The main trade-off is setup and oversight: the value depends on how much context you feed it and how carefully you define what it may do unattended.

How do you use Vellum?

  1. 1Create your Vellum account
    Sign up on the Vellum site to start. The assistant is free to download and begin with, and cloud plans can be added later if you want hosted compute.
    Vellum — Create your Vellum account
  2. 2Choose local or cloud hosting
    Decide whether the assistant runs on your own machine or in a Vellum Cloud container. Local keeps memory and credentials on your device by default.
  3. 3Connect your tools
    Enable the built-in skills you need, such as Google Workspace, Slack, Notion and filesystem access, then add community plugins or custom skills as required.
  4. 4Set permissions for sensitive actions
    Review the permission prompts for tasks like emailing new contacts or editing important files. Mark recurring, low-risk actions as trusted only after you are comfortable with them.
  5. 5Teach the assistant about your context
    Share the people, projects and preferences you want remembered. Memory carries across conversations, so the assistant becomes more useful as context accumulates.
  6. 6Review plans and credits
    Compare the Mighty, Super and Ultra cloud plans on the pricing page if you need dedicated compute, storage and bundled credits.
    Vellum — Review plans and credits

Pros and cons

Pros

  • Long-term memory stores people, preferences and work context across every conversation.AI
  • Local-first by default: memory and credentials stay on your machine unless you opt into cloud.AI
  • Model-agnostic, with switching between OpenAI, Anthropic, Gemini and open-weight models via Ollama.AI
  • Fine-grained permissions gate risky actions until you mark them as trusted.AI
  • Extensible through community plugins and custom skills on top of built-in Google Workspace, Slack and Notion skills.AI

Cons

  • Real value depends on upfront setup: integrations, permissions and context need configuration before the assistant is useful.AI
  • Cloud plans start at $30 per month, which is a meaningful cost on top of any model API spend.AI
  • Local-first operation puts backup, sync and security responsibility on the user.AI
  • The plugin and skills ecosystem is community-driven, so quality and coverage vary.AI
  • Heavy reliance on memory means accuracy is only as good as what the assistant has been told, and stale or wrong context can persist.AI

How much does Vellum cost?

Free trial

The assistant is free to download and start, so you can evaluate core functionality before committing to a paid cloud plan. The homepage does not describe a separate time-limited trial.

Learn more

Pricing

Vellum is free to download and start. Cloud subscriptions are priced at $30 per month (Mighty), $100 per month (Super) and $200 per month (Ultra), each with dedicated compute, storage and bundled credits. The pricing page lists the full comparison of every plan.

Learn more

Integrations

Built-in skills cover Google Workspace, Slack, Notion and the local filesystem. Additional capabilities come through community plugins and custom skills, and models can be routed through OpenAI, Anthropic, Gemini or open weights via Ollama.

Features

Core features include persistent long-term memory, autonomous task execution, a granular permissions system, model switching across OpenAI, Anthropic, Gemini and Ollama, built-in skills for Google Workspace, Slack, Notion and the filesystem, community plugins, custom skills, and access from Mac, iOS, Android, web, voice, email, Telegram, Slack and terminal.

Frequently asked questions about Vellum

  • How much does Vellum cost?
    Vellum is free to download and start. Cloud subscriptions are priced at $30 per month (Mighty), $100 per month (Super) and $200 per month (Ultra), each with dedicated compute, storage and bundled credits. The pricing page lists the full comparison of every plan.
  • Does Vellum offer a free trial?
    The assistant is free to download and start, so you can evaluate core functionality before committing to a paid cloud plan. The homepage does not describe a separate time-limited trial.
  • How do you use Vellum?
    The walkthrough on this page covers 6 steps: 1. Create your Vellum account 2. Choose local or cloud hosting 3. Connect your tools 4. Set permissions for sensitive actions 5. Teach the assistant about your context 6. Review plans and credits.
  • What platforms does Vellum support?
    Vellum is available on MacOS, Web App, iOS and Android.
  • What does Vellum integrate with?
    Built-in skills cover Google Workspace, Slack, Notion and the local filesystem. Additional capabilities come through community plugins and custom skills, and models can be routed through OpenAI, Anthropic, Gemini or open weights via Ollama.
  • What are the limitations of Vellum?
    Real value depends on upfront setup: integrations, permissions and context need configuration before the assistant is useful. Cloud plans start at $30 per month, which is a meaningful cost on top of any model API spend. Local-first operation puts backup, sync and security responsibility on the user.

Status

StatusActive
Views0
Outbound clicks0
Added10/21/2026

Platforms

MacOSWeb AppiOSAndroid

Pricing

FreemiumSubscription

Categories

AI Chat BotProactive AI AssistantTask CompletionChatLife AssistanceProductivity