</> Open Tag

Model-agnostic Claude Tag

Tag it in Slack.
Run any model.

Open Tag is an AI teammate you @mention in the channel. It does the job and posts the result — on Claude, GPT, Gemini, or the open-source model you host yourself.

Get started Trust model
Slack · Your key · Open-source models welcome

Quick start

Same path as tagging a person. About two minutes, then a real job.

  1. 1

    Invite Open Tag to a channel

    It only reads channels a human invites it into. Remove it and access ends with the membership. No workspace-wide ingest.

  2. 2

    Point it at a model

    Bring your own key — Anthropic, OpenAI, Google, or an open-source model on Ollama, vLLM, or your inference box. We do not pick a model for you, and we do not mark up tokens.

  3. 3

    Tag it with a job

    Ask for work, not a tutorial. It plans, runs, waits if something is irreversible, and posts the artifact back in the thread.

Stuck? Email support@opentag.bot with your workspace name.

What it does

A chatbot tells you how. Open Tag finishes the work in the channel.

Finished work, not advice

Answers, files, diffs, and updates in your connected tools — with the steps behind them, so you can check the work.

Sourced answers

Ask why something was decided and get the thread, the doc, and who said it — not a search dump.

The model you actually want

Frontier APIs when you need them. Llama, Qwen, Mistral, DeepSeek when you want cheaper, faster, or fully local.

Tools you already have

Connect the apps the job needs. Permissions follow the person who asked, not a shared master key.

Repeatable jobs

When the same request keeps showing up, promote it to a standing run instead of retyping it every Monday.

A stop button

It shows the plan so you can halt it. Sending, spending, merging, or deleting waits on a person.

Works with what you bring

OpenClaw-shaped: your models, your chat, your keys. Slack is where it lives today.

Models

  • Claude
  • GPT
  • Gemini
  • Llama
  • Qwen
  • Mistral
  • DeepSeek
  • Ollama
  • vLLM
  • Your endpoint

Where you talk

  • Slack channels
  • Threads
  • Direct messages

Point it at an open-source model

Llama, Qwen, Mistral, DeepSeek — served by your provider or running on your own hardware. Cheaper, faster, and more reliable than depending on a single closed vendor.

Cheaper

Open weights run on commodity inference at a fraction of frontier per-token pricing, and you pay your provider directly with no markup from us.

Faster

Smaller open models answer in a fraction of the time, and hosting them close to your data cuts the round trip that makes a Slack thread feel slow.

More reliable

Pin a version that never changes underneath you, sidestep vendor rate limits and deprecations, and fail over to another model without rewriting anything.

Three guarantees

Taken from how we already talk about trust — made explicit here, not buried in a policy PDF.

Invited channels only

Open Tag reads a channel only after a human invites it in. There is no workspace-wide scrape.

Irreversible waits

Sending, spending, merging, or deleting stops at an approval. The gate is in the runtime, not in the model's mood.

Your access, your key

Integrations follow the person who connected them. The model is the one you configured. We do not train on your private content.

Full trust model →

Next to Claude Tag

Same idea — tag a teammate in Slack — without being locked to one lab's model.

Claude Tag Open Tag
Where it lives Slack Slack
How you talk to it Tag it in a channel Tag it in a channel
Model Anthropic Any model you bring, including open-source
Keys Vendor-managed Bring your own. Local weights stay on your hardware
Channel access What you install it into Only channels a human invites it to

FAQ

What is Open Tag, exactly?

An AI teammate you tag into Slack. It has the context you give it in that channel, talks to the model you configured, and posts finished work: answers, files, and changes in connected tools.

How is this different from a chatbot?

A chatbot tells you how. Open Tag does the job and leaves an artifact in the thread. You can still stop it, and anything irreversible waits for a person.

Does it read all our Slack messages?

No. It reads a channel only after you invite it in. Remove it and access ends with the membership.

Can it make mistakes?

Yes. It can be wrong. It shows its plan so you can stop it, cites sources so you can check it, and asks before anything leaves your company that you cannot undo.

What happens to our data?

Content needed to run a job is processed so the teammate can finish it. We do not sell it, and we do not use your private content to train our own models. Your model provider — or your own box, if you run open weights locally — sees what you send it. Details are on Privacy and Security.

Is this Gini?

No. Gini is a hosted coworker with its own cloud machine and usage-based plans. Open Tag is the model-agnostic, tag-in-Slack version of that idea: your model, your key, Slack as the desk.

Please note