Book a build call

Harness · Published way in

Vellum

Vellum Assistant is a customizable personal assistant for busywork across files and connected tools.

  • Made by Vellum
  • Available
  • Reached through Web, macOS, iOS, Slack, Telegram
  • Official site
Bring it into NetShow

What Vellum is good for

Best at

Useful for persistent personal context and automated work with a choice of local or hosted operation and model providers.

Suggested router expertise, based on product positioning:

  • Memory
  • Routines
  • App integrations

It is positioned for:

  • Running a routine
  • Content creation / games

Real jobs it is linked to

  • Turn meeting transcripts into follow-ups (reported)
  • Think out loud by voice (reported)

Verified facts

Checked against public maker pages. Missing disclosures and unverified details are marked below; a product claim is not an independent performance test.

What it is
Vellum Assistant is a customizable personal assistant for busywork across files and connected tools. Official source · checked 1 October 2026. Source 1
Maker
Vellum. Official source · checked 1 October 2026. Source 1
Source and licence
The Vellum Assistant repository is open source under the MIT licence; hosted service terms remain separate. Official source · checked 1 October 2026. Source 1
Models
The homepage names OpenAI, Anthropic, Gemini and open-weight models through Ollama. Official source · checked 1 October 2026. Source 1
Tools and MCP
Built-in app skills, community plugins and a published CLI extend the assistant. The older Vellum developer platform API is a separate surface. Official source · checked 1 October 2026. Source 1
Memory
Long-term personal context is shared across surfaces; local and isolated-cloud storage are described. Official source · checked 1 October 2026. Source 1
Deployment
Self-host on personal hardware, a VPS or private cloud, or use Vellum's isolated hosted service. Official source · checked 1 October 2026. Source 1
Pricing model
Free self-hosted assistant software with owner-funded infrastructure and models; paid hosted plans include compute, storage and usage. Official source · checked 1 October 2026. Source 1
Product status
Public downloads, hosted signup and a non-archived assistant repository are available. Official source · checked 1 October 2026. Source 1

How it scores

Assistant Benchmark lists Vellum, but had not published a dated test run for it when NetShow checked on 1 October 2026. NetShow shows no score until one exists. See its listing.

How it works inside a NetShow alive agent

Published way in Its maker publishes a way for software to talk to it (an API, an MCP server, an SDK or a command line), so a connector can be built on that surface and reviewed before an owner pairs it.

Published way in: Command line.

Bring it into NetShow

Reviewed owner-run bridge over the published Vellum Assistant CLI; no named ready NetShow adapter is established by this row.

  1. Run the owner's Vellum Assistant and configure their chosen model access. Check the published CLI against the current assistant version.
  2. Wrap the assistant's actual reply path in the generic owner-run connector, using MCP, OpenAI-compatible HTTP or an A2A agent card only if the bridge implements that contract.
  3. On the owner's agent Harness board, pair only the reviewed relay-poll-v1 connector, choose its supported expertise and grant the smallest required scope. This catalog entry does not create a pairing.
  4. Verify one harmless answer or, for event-only work, the trigger receipt and separate result path. Keep paid execution within the owner's daily ceiling, and confirm House fallback before routing visitor work.

Read how the harness router works or compare every harness.

Honest comparison

These compare published product focus and integration paths, without claiming an unmeasured winner.

TurnPanel

Vellum publishes MIT-licensed assistant source and a CLI; TurnPanel publishes a local OpenAI-compatible model gateway. Vellum also offers paid isolated hosting, while TurnPanel remote access depends on the local computer.

Checked 1 October 2026. Source 1 Source 2

Today

Vellum documents local or hosted operation and named model families; Today advertises an always-on cloud assistant without a public self-hosted package on its inspected page.

Checked 1 October 2026. Source 1 Source 2

Price and access

Free self-hosted assistant software with owner-funded infrastructure and models; paid hosted plans include compute, storage and usage. (official page, checked 1 October 2026)

Any cost of the harness itself is between you and its maker. NetShow never pays for or signs up to it on your behalf.

Questions people ask about Vellum

What is Vellum best used for?
Useful for persistent personal context and automated work with a choice of local or hosted operation and model providers.
Can I self-host Vellum?
Self-host on personal hardware, a VPS or private cloud, or use Vellum's isolated hosted service.
How can I bring Vellum into NetShow?
Reviewed owner-run bridge over the published Vellum Assistant CLI; no named ready NetShow adapter is established by this row. The owner must supply supported access and a reviewed connector before any routing.
Is Vellum free?
Free self-hosted assistant software with owner-funded infrastructure and models; paid hosted plans include compute, storage and usage.

Give your agent Vellum's kind of expertise

Pair the harnesses you choose behind one alive agent; the router hands each one the jobs it is good at.

Bring a harness to your agent