Back to Gap Map Product / UX / Orchestration & agents

Vellum

open source / Overall score: 2.7

Personal AI assistant across web, macOS, iOS and a CLI, with episodic, semantic and procedural memory, that takes actions on the user's behalf across channels including Slack and Telegram. It runs proactive check-ins and isolates credentials out of process. The company pivoted to it in 2026 from an LLMops platform, which remains available.

Openness

5 high confidence

5.0

license: MIT(OSI, Copyright 2025 Vellum AI)
source: public(the full agent - gateway, clients, skill framework, memory - in github.com/vellum-ai/vellum-assistant)
core features withheld: no — README: managed runtime on Vellum Platform or self-hosted, "Same codebase, same data model"
self-host: yes(by design)
managed-cloud: hosting convenience, not a feature tier

Vellum's complete agent — the gateway, its Mac, Slack and Telegram clients, the skill framework and the memory system — is published under MIT and can be self-hosted by design. The vendor's own README states that its managed cloud runtime and the self-hosted version share the same codebase and data model, so the hosted option is a convenience rather than a separate, feature-gated tier.

https://github.com/vellum-ai/vellum-assistant recorded 2026-08-13
PRIMARY - full open source agent source (gateway, clients, skill framework, memory) under an MIT license badge, 1.1k stars, 156 forks
https://github.com/vellum-ai/vellum-assistant/blob/main/LICENSE recorded 2026-08-13
PRIMARY - MIT License text, 'Copyright (c) 2025 Vellum AI'
https://raw.githubusercontent.com/vellum-ai/vellum-assistant/main/LICENSE recorded 2026-09-29
The LICENSE file is the standard MIT text: "MIT License" with "Copyright (c) 2025 Vellum AI".
https://raw.githubusercontent.com/vellum-ai/vellum-assistant/main/README.md recorded 2026-09-29
The README states the license and hosting model: "Vellum Assistant is open-source software built by" Vellum AI, "Free to use and modify under MIT", and "Managed runtime on Vellum Platform, or self-hosted. Same codebase, same data model." Searched for enterprise, ee/, license key: none.

Adoption

2 low confidence

2.0

Vellum discloses no download, user or usage metric, and the assistant is a self-hosted app rather than a registry package, so GitHub stars are the only available proxy, used here strictly as a last resort. That places it as a small, newly launched entrant.

https://github.com/vellum-ai/vellum-assistant recorded 2026-08-13
1.1k stars, 156 forks (last-resort star proxy; no other adoption signal disclosed)
https://www.vellum.ai recorded 2026-08-13
landing page with no disclosed user or usage numbers

Capability

3 low confidence

3.0

Solid personal-agent feature breadth, all of it visible in the open MIT source: 8-type memory, real-world action through managed OAuth connections, multi-channel reach across Mac, web, iOS, Slack, Telegram and phone, out-of-process credential isolation and an ambient proactive engine. No benchmark and no frontier differentiator, so this is a mid-tier judgment on features.

https://raw.githubusercontent.com/vellum-ai/vellum-assistant/main/README.md recorded 2026-09-29
The README describes the agent with "8 different types of memory" and channels "macOS, Windows, iOS, Web, Voice, Email, Telegram, Slack, Twilio", plus "Credentials live in a separate process and never reach the model."
https://www.vellum.ai recorded 2026-09-29
The homepage structured data lists "Open source (MIT license)", "Proactivity engine with hourly check-ins" and "Permission and credential isolation"; no benchmark figure found.

Verified 2026-09-27