# Vellum vs Gumball
> Personal AI assistant vs Gumloop's workplace sidekick
> Last updated: September 11, 2026
> Comparison: https://www.vellum.ai/compare-assistants/vellum-vs-gumball

## Overview
Vellum is an open-source personal AI assistant with free cloud hosting that runs on your computer or in Vellum Cloud, with continuous structured memory across Mac, iOS, Android, web app, voice, email, Telegram, and Slack. Gumball is Gumloop's cloud-native workplace AI sidekick, integrated into Slack, email, and the Gumloop automation canvas with daily catch-up digests (Daily Chew), Smart Inbox triage, and automated Meeting Prep. While both systems automate complex workflows across connected apps, Vellum prioritizes true personal ownership, local machine execution, and lifetime memory, whereas Gumball operates as a cloud-based team automation agent tied to monthly execution credits.

## The contenders
### Vellum
Website: https://vellum.ai

Vellum is a voice-first personal AI assistant that learns how you work and can improve itself around your habits. Its companion works alongside you in real time, or takes on longer tasks in the background. It follows you across devices and apps with one shared memory, and can run in Vellum Cloud or on your own machine.

Why choose Vellum:
- Model agnostic: Use any frontier AI model, open or proprietary, and switch anytime.
- Truly yours: Give it a name and personality. It learns you, your work, and your tools.
- Self-improving: Gets better the more you use it, with one memory across every device.
- Private by design: Credentials stay secure, and you choose where your data lives, including self-hosting.
- Built for busy business owners: Offload work and make more time for life.

### Gumball
Website: https://www.gumloop.com

Gumball is Gumloop's personal AI sidekick, created automatically for every Gumloop account. It connects to your SaaS apps, monitors communications with Daily Chew morning summaries, triages email via Smart Inbox, briefs meetings, and executes multi-agent workflows in the cloud.

Why choose Gumball:
- Native Daily Chew morning catch-up digests and automated pre-meeting briefs
- Deep integration with Gumloop's no-code workflow canvas and team-shared skills
- Smart Inbox automated categorization and draft replies for Gmail and Outlook
- Cloud-native background monitoring across 100+ business connectors and MCP integrations

## The scorecard

### channels

Vellum: 9 surfaces, one memory
iOS app with Home Screen widgets, native Mac app, web app, voice, email, Slack, Telegram, Microsoft Teams, and Discord, now an official channel with a setup wizard. Sent messages can be edited in place in Slack and Telegram. Every surface shares the same memory and identity.
Vellum runs natively on Mac and iOS with Home Screen widgets, plus web app, voice, email, Slack, Telegram, Microsoft Teams, and Discord. Sent messages can be edited in place in Slack and Telegram. Every surface shares the same memory, so you can start a task on your Mac and pick it up on your phone or in a chat.

Gumball: Slack DMs, email, and web app interface
Direct messages in Slack, dedicated gumball@gumloop.ai email address, and web dashboard.
Users interact with Gumball by direct messaging the Gumloop Slack app, emailing requests to gumball@gumloop.ai, or chatting within the Gumloop web application. Mobile access is available through responsive mobile web and Slack mobile.

### execution

Vellum: One sentence to real output
Build me a habit tracker gets you a working, interactive app in the conversation. Read this page uses the browser extension to read your open tabs and act on the page you are looking at. Call the dentist and reschedule places an actual phone call with its own voice.

Gumball: Cloud node execution and headless browser automation
Executes API calls, web scraping, and Python scripts across distributed cloud workers.
Gumball executes actions through Gumloop's distributed worker nodes, performing web scraping, data transformation, LLM synthesis, and code execution in secure cloud sandboxes without consuming local computer processing.

### hosting

Vellum: On-device, managed cloud, or self-hosted
Run on-device as a native Mac app, on Vellum's managed cloud, or self-host on your own VPS.
Vellum gives you the full range: run the native Mac app entirely on your own hardware, use Vellum Cloud for managed uptime and cross-device sync, or self-host on your own VPS for full infrastructure control.

Gumball: Fully managed cloud infrastructure
Runs on Gumloop cloud workers with scheduled background triggers; no local hosting.
Gumball runs 24/7 on Gumloop's cloud worker infrastructure, executing cron jobs, morning digests, and event-based triggers even when user laptops are closed. There is no native local desktop runner or offline mode.

### integrations

Vellum: 50+ managed OAuth integrations
Native agent skills plus 1-click integrations with 50+ services through managed OAuth, including native Calendly and Eventbrite. Your assistant acts in real services without you touching tokens or API keys.
Vellum connects to 50+ services through managed OAuth: Google, Slack, Notion, GitHub, Linear, Calendly, Eventbrite, and more. Your assistant acts in real services on your behalf without you handling tokens or API keys.

Gumball: 100+ business connectors and MCP integrations
Extensive pre-built connectors for Gmail, Outlook, Slack, Salesforce, HubSpot, Linear, and MCP servers.
Gumball connects to over 100 enterprise SaaS tools and standard Model Context Protocol (MCP) servers, enabling deep actions across customer relationship management, project tracking, cloud storage, code repositories, and data warehouses like Snowflake.

### memory

Vellum: Structured memory that builds itself, everywhere
Vellum builds a structured memory of your projects, people, and preferences automatically, across every surface. It sharpens with every conversation and correction, and there is nothing to configure: no files to curate, no engine to pick, no embedding provider to wire up.
Vellum builds a structured memory graph of your projects, people, and preferences across every surface. It sharpens over time through use and explicit corrections, so each conversation picks up where the last left off.

Gumball: Tone, design guidelines, and situation tracking
Tracks personalized tone and design rules, situational highlights, and past run feedback in cloud storage.
Gumball curates personalization settings for tone and design, tracking user feedback across Daily Chew runs, meeting briefs, and team-shared skills within Gumloop cloud storage. Context and situations are organized around workspace tasks and connector runs.

### modularity

Vellum: Adapts to you personally
Builds its own skills, learns your tone, and adapts based on what you reward and correct. Install new skills from a growing library, or just ask and it figures out a way.
Vellum adapts at the individual level: it learns your communication style, builds custom skills from your instructions, and updates its behavior based on your feedback. Each assistant is shaped by the specific person using it.

Gumball: Drag-and-drop workflow canvas and team-shared skills
Extensible visual node canvas, reusable SKILL.md imports, and multi-agent delegation.
Gumball leverages Gumloop's modular agent platform: users can chain AI nodes on a drag-and-drop canvas, import GitHub repositories containing SKILL.md playbooks, and delegate specialized assignments to subagents or team-shared department agents.

### open-source

Vellum: Open source under MIT
Open source under MIT license. Full codebase available on GitHub.
Vellum is open source under the MIT license. The full codebase is on GitHub and can be self-hosted, audited, and extended.

Gumball: Proprietary SaaS platform
Proprietary cloud platform with closed-source engine, web builder, and orchestration harness.
Gumball and the Gumloop agent orchestration engine are closed-source proprietary software built by AgentHub Inc. The underlying runner, skill execution pipeline, and canvas code cannot be inspected, self-hosted, or independently modified.

### pricing

Vellum: Free Base, paid from $30/mo
Free Base plan with free credits to start. Paid plans from $30/mo (Mighty), $100 (Super), and $200 (Ultra) with pay-as-you-go credits, configurable compute and storage, and your assistant's own email and subdomain on higher tiers.
Vellum offers a free Base plan with credits to start. Paid plans run from $30/mo (Mighty) to $100 (Super) and $200 (Ultra) with pay-as-you-go credits, configurable compute and storage, and a personal assistant email and subdomain on higher tiers.

Gumball: Credit-based usage pricing with 14-day free trial
14-day free trial, Pro plan from $37 per month on a credit slider, and custom Enterprise contracts.
Gumloop operates on a usage-tiered credit model. Users start with a 14-day free trial of the Pro plan, which begins at $37 per month for 20,000 credits (sliding up to $1,840 for 1M credits) with unlimited team seats and credit overages. Enterprise plans require custom sales agreements.

### privacy

Vellum: Local or cloud, always yours
Run locally on Mac, in Vellum Cloud, or self-host. Vellum never has access to your data on any deployment path.
Vellum gives you three deployment paths: native Mac app on your own hardware, Vellum Cloud with per-customer isolation, or self-hosted on your own infrastructure. On every path, Vellum never has access to your data.

Gumball: Multi-tenant cloud with optional VPC deployment
Processes data across Gumloop cloud infrastructure, with dedicated VPC options on Enterprise contracts.
Gumball executes in Gumloop's multi-tenant cloud environment, reading connected SaaS inboxes, calendars, and CRMs. Enterprise plans offer dedicated virtual private cloud (VPC) deployments in AWS, Azure, or Google Cloud so organizational data remains inside company boundaries.

### schedules

Vellum: Built-in scheduling, always running
Built-in scheduling for daily briefings, weekly reports, and hourly checks that just run on a cadence.
Vellum has native scheduling for recurring tasks: daily briefings, weekly reports, hourly checks. Scheduled tasks run in the cloud whether or not you have a session open.

Gumball: Scheduled Daily Chew digests and event-based triggers
Automated morning briefings, pre-meeting triggers, and custom interval schedules.
Gumball runs automated background schedules including the Daily Chew morning newsletter, calendar-triggered meeting briefings, and custom cron triggers set up on the Gumloop workflow canvas to monitor third-party web changes and app events.

### security

Vellum: Isolated credential vault
Credentials live in an isolated vault and never enter the model layer, so prompt injection can't exfiltrate your keys. Every sensitive action asks permission with a risk badge before it runs, and guardian approvals land in one inbox.
Vellum stores every connected service credential in an isolated vault that never touches the LLM inference layer. Prompt injection attacks cannot reach your API keys or OAuth tokens.

Gumball: Role-based access, connector permissions, and credit gates
Admin-governed connector controls, role-based skill permissions, and credit rate limits.
Gumball relies on Gumloop's enterprise governance controls: organization admins can restrict integrations, enforce role-based access for skill creation, and monitor model usage. API credentials are encrypted in cloud storage, and credit usage tracks live per interaction.

### setup

Vellum: Sign up and start talking
No terminal, no API keys, no daemon. Voice-first onboarding is the default: your first experience is speaking, not configuring. Free, no credit card, up in minutes.
Vellum setup is sign up and start talking. Create an account and your first sentence gets real output. Voice is the default first experience. No terminal, no config files, no API keys, no infrastructure to provision.

Gumball: Instant account provisioning with OAuth connectors
Auto-created on account signup; connects to workplace apps via standard browser OAuth.
Gumball is automatically provisioned the first time a user signs into Gumloop. Connecting email, calendar, and Slack requires only clicking standard OAuth authorization buttons, with no local packages, dependencies, or server setup required.

### storage

Vellum: Storage
Vellum Cloud: 3 GB RAM and 4 GB storage by default. Mac app runs on standard Mac hardware.

Gumball: Cloud artifacts and team knowledge base
Stores generated artifacts, interactive apps, and company brain documents in Gumloop cloud.
Files, dashboards, generated web applications, and documents produced by Gumball are stored as cloud Artifacts with shareable URLs and access controls. Knowledge is indexed in Gumloop's Company Brain in the cloud.

### trust

Vellum: Permissions you can see
Every sensitive action asks first with a risk badge, guardian approvals land in one inbox, and you can export or delete your data at any time. The trust model is written down in the docs.

Gumball: Review gates and interactive Slack feedback buttons
Interactive Needs Work feedback in Slack and step-by-step plan reviews before agent execution.
Gumball incorporates human-in-the-loop review points: Daily Chew situations provide step-by-step action plans users can inspect before sending to chat, and Slack briefs feature interactive Needs Work buttons to correct agent context directly in-thread.

## Where they split

### Which one is right for you?
Gumball is Gumloop's cloud sidekick for workplace operations, automating morning digests, meeting prep, and inbox triage across corporate SaaS tools. It ties directly into team automation canvases, with your data, memory, and execution metered by monthly cloud credits.
Vellum handles those same workplace briefs, schedules, and daily tasks, but lives on your computer with complete data ownership. You get an autonomous personal assistant out of the box with zero technical setup, continuous memory across your devices, and your private data never trapped in a closed cloud.

### Which setup fits you best?
Gumball operates across the Gumloop web application, Slack direct messages, and dedicated email threads, connecting directly to cloud SaaS APIs and automated workflows.
Vellum lives where you work: a native Mac app that controls your local computer, mobile apps on iOS and Android, web app, voice calling, email, Slack, Telegram, and terminal, all connected to one persistent assistant.

### How do you want privacy and security handled?
Gumball processes data in Gumloop's multi-tenant cloud with role-based admin controls, reserving private VPC deployments for Enterprise contract customers.
Vellum gives you complete architectural ownership: run locally on your physical device with zero cloud transmission, self-host under MIT, or use Vellum Cloud with isolated credential vaults that keep your keys and passwords private.

## Lingering questions

### Should I choose Vellum or Gumball for a personal AI assistant?
Choose Vellum if you want an autonomous personal assistant for daily work, errands, and private projects with full data sovereignty, free cloud hosting, and continuous memory that lives on your Mac, phone, and messaging apps. Choose Gumball if your primary goal is automating team SaaS workflows inside Slack, email, and Gumloop's visual automation canvas.

### When does Gumball make more sense than Vellum?
Gumball makes sense if your team already relies on Gumloop for visual workflow automation and you want a shared AI sidekick that handles morning inbox triage, pre-meeting briefs, and connector-driven tasks inside your corporate Slack workspace. If you want a personal assistant that manages your local files, runs natively on your computer, and keeps your private context independent from corporate cloud meters, Vellum is built for you.

### How much does Vellum cost compared with Gumball?
Vellum includes free cloud hosting when you sign up. Paid plans upgrade compute and storage: Mighty at $30 per month, Super at $100 per month plus a $10 platform fee, Ultra at $200 per month plus a $10 platform fee, and Custom. Self-hosting has no platform fee. Gumloop offers a 14-day free trial, with paid Pro plans starting at $37 per month for 20,000 credits on a usage slider, scaling up to custom Enterprise contracts.

### How does Gumball's security compare to Vellum's credential vault?
Gumball secures credentials in Gumloop's cloud repository with role-based admin controls and optional VPC deployment for Enterprise clients. Vellum isolates credentials in a dedicated local or cloud execution vault where passwords and API keys never enter the prompt context, while keeping the entire assistant architecture open-source and auditable.

### Can I self-host Gumball or run it on my own computer?
No. Gumball runs exclusively on Gumloop's managed cloud infrastructure, with private VPC deployment available only for Enterprise contract customers. Vellum can run directly on your own computer as a native Mac app with local memory, self-hosted on your private server under the MIT license, or deployed in managed Vellum Cloud.

### Can Vellum brief me on meetings and summarize my inbox like Gumball?
Yes. Vellum connects to your Google Calendar, Outlook, and email accounts to deliver morning briefings, summarize incoming messages, draft replies in your personal tone, and prepare agenda briefs before upcoming meetings, all while keeping your data under your own control.

### What messaging apps do Vellum and Gumball support?
Gumball supports direct messaging inside Slack, a dedicated gumball@gumloop.ai email address, and the Gumloop web application. Vellum connects across Mac, iOS, Android, web app, voice calling, email, Telegram, Slack, and terminal, sharing continuous memory across all surfaces.

### Where does my data live with Gumball vs Vellum?
With Gumball, all prompt history, connector tokens, generated artifacts, and meeting notes are stored in Gumloop's cloud database. With Vellum, you choose your data residency: keep your memory entirely local on your physical Mac, self-host on your own infrastructure, or use managed Vellum Cloud with isolated credential vaults.

### Can I use models like Claude, GPT, or local models with Gumball?
Gumball automatically selects models based on task complexity from the models permitted by your organization. Vellum gives you explicit multi-model routing control, letting you select Anthropic Claude, OpenAI, Google Gemini, OpenRouter, or private local models running offline via Ollama.

### Can Gumball manage files on my physical laptop?
No. Gumball operates strictly across cloud SaaS APIs and web connectors. It cannot directly read, edit, or organize files on your personal physical computer. Vellum runs natively on your machine, managing local folders, running terminal commands, and coordinating desktop workflows directly.

### Is Vellum or Gumball easier to use for non-technical users?
Both provide zero-friction onboarding. Gumball is ready on signup via web and Slack, while Vellum provides a complete personal assistant across Mac, phone, web, voice, email, and messaging with no server configuration needed. The difference is architectural: Gumball is a cloud automation teammate metered by credits, while Vellum is your personal assistant that you own.

---
Full comparison index: https://www.vellum.ai/compare-assistants
Try Vellum: https://www.vellum.ai