attest_tag
Sign inTry it now
Menu
Comparisons

Comparisontheir pages read 22 Sep 2026

Runbear vs attest_tag

Runbear sells “no engineering” to ops teams — its homepage’s words — and attest_tag is written for the engineer. The two bullets Runbear lists under Business at $399 a month, a model key of your own and an audit log of every prompt and tool call, are in the MIT binary you can run yourself; the audit log, approval tiers and budgets are in the free plan too, and every write is held for a Confirm in the thread by default.

Try it nowDeploy it yourselfSkip to the table
The short versionTheir page, oursSide by sideQuestionsTry both

Which model answers

Any endpoint, per thread

What a turn costs, and who sees it

Model and cost under every reply

Where it runs

Hosted, or one binary of yours

Who holds the credentials

The proxy. Never the model

Who may approve

Never your own request

What you can audit

Every call, exportable

Pick attest_tag when

  • A credential is injected by an in-process proxy one hop before the request leaves, and scrubbed from every response before it reaches the model or Slack — the model only ever emits a URL and a method.
  • A monthly budget for the workspace, another per channel, and a per-user hourly rate limit; the bot pauses at the ceiling and says so, rather than reporting having crossed it.
  • Nobody approves their own request — enforced three times over, once in SQL — and a request routes to the least powerful tier that covers every call in it.
  • A fix job runs in a separate worker container that clones the branch, runs the tests before and after the change, and opens a draft pull request it never merges.
  • A flat monthly fee by plan size, never per seat, and every feature in every plan — approval tiers, audit export, budgets and two-factor sign-in are in the free plan and the open-source binary.
  • You need HIPAA: run the same binary on your own infrastructure so protected health information never leaves it, and we sign a business associate agreement with you on request.
The ten questions, side by side

Their page, ours

What the hero says, and what the rest of the site does not.

A hero line is a promise about the product. The questions an engineer asks before rolling one out are usually answered somewhere else on the site — or not answered. These are the ones we looked for.

Runbeartheir hero, verbatim

“AI seats sit empty. AI teammates show up.”

Runbear puts AI teammates into Slack, Microsoft Teams and Discord channels, where they answer @mentions and direct messages and act through connected tools. It is priced by credits rather than seats.

What their pages do not say

  • Which model ran, shown under the reply
  • What a turn cost, shown under the reply in the thread
  • Where the OAuth tokens and API keys live — not on the pages that rendered for us
  • A Confirm before every write by default, not only the sensitive ones
  • Whether you can run it yourself, or read the source
  • On Team at $99 a month, your own model key or an audit log — both are Business bullets at $399 a month

On the public pages listed under the table, as of 22 Sep 2026. An absence is a fact about a page, not about the product; if one of these has since been answered, we will change it.

attest_tagthe same questions, in writing

Side by side

Ten questions worth asking either way.

Our column is the same on every comparison, and each cell ends where the mechanism is shown. Theirs is Runbear’s own pages, quoted or closely paraphrased, as read 22 Sep 2026.

01Which model answers

attest_tag

Any OpenAI-compatible endpoint — OpenRouter, a vendor API, or a model you host. Ships on google/gemini-flash, with a heavy model routed to automatically for code and long threads, and a per-thread override.
Models and cost

Runbear

“Provider chosen per agent; the credit rate follows the model you pick” — Claude, GPT or Gemini, on Runbear's access or “Bring your own LLM key (BYOK)”, a Business-tier bullet at $399 a month. Their Traces API returns “the model, latency, input/output, intermediate steps” for each run; whether the model that answered is shown under the reply in Slack is not stated.
02What a turn costs, and who sees it

attest_tag

Metered per completion and printed under the reply beside the model that answered, with !usage for the month by channel. A monthly budget for the workspace, another per channel, and a per-user hourly rate limit — the bot pauses at the ceiling rather than reporting having crossed it.
Audit and budgets

Runbear

“A credit reflects the model work a task actually uses.” Their docs give ranges — “Question answered from knowledge: 10 to 20 credits”, “Task involving several tool calls: 30 to 50 or more credits” — a billing page that “shows overall consumption”, a Usage page per agent, and a Credits API that returns “Every metered call”. A cost shown under the reply, in the thread, is not stated.
03Where it runs

attest_tag

Hosted by us, or one MIT-licensed Go binary with the console embedded on your own infrastructure — Docker, Cloud Run, Fargate, Container Apps or Kubernetes. The same code either way. Where HIPAA applies, run it yourself so nothing leaves your network, and we sign a business associate agreement with you.
Self-hosted

Runbear

Not stated on their site.
04Who holds the credentials

attest_tag

The service, never the model. Sealed with AES-256-GCM at rest — under a key you hold, on a self-host — injected by a proxy at the network edge, and scrubbed back out of every response before it reaches the model or the thread.
Credentials, in the guardrails

Runbear

“Per-user auth on shared agents”, a Business-tier bullet; the Enterprise page adds “inherits asker's perms”. Where the OAuth tokens and API keys are stored, how they are encrypted and who can see them is not stated on the pages we could read; their pricing FAQ answers and Trust Center did not render when we checked.
05Before it changes something

attest_tag

Any method other than GET, HEAD or OPTIONS is held for a card in the thread — Confirm, Cancel, or say what you actually wanted. Pending writes expire after five minutes. A fix job opens a draft pull request and never merges.
Human-in-the-loop writes

Runbear

“Set approval gates on sensitive actions — refunds, deletes, sends”, on the Enterprise page, and “The agent drafts, a named human signs off”. The homepage: “Anyone in the channel can send it or edit first”. A Confirm on every write is not stated on their site.
06Who may approve

attest_tag

Writes that grant access leave the thread and go to an approver’s DM. Nobody approves their own request — enforced three times over, once in SQL — and a request routes to the least powerful tier that covers every call in it.
Access requests and tiers

Runbear

“The agent drafts, a named human signs off.” Which roles may sign off, and whether a requester may approve their own action, is not stated on their site.
07What you can audit

attest_tag

Every turn, tool call, proxied request and completion — with tokens and cost — in a console you run, exportable as CSV. Self-hosted, it is a table in your own database.
The audit trail

Runbear

“Audit log of every prompt, tool call, and reply”, listed under Business at $399 a month; the Enterprise page adds “SIEM-exportable”. An admin audit log, described in their docs as part of the enterprise entitlement, “records configuration changes, not conversations”.
08Source and licence

attest_tag

MIT. The repository is the product: the hosted service runs the same binary you can clone, read and build.
The same binary, inside your network

Runbear

Not stated on their site.
09What it costs

attest_tag

A free plan that does not expire — $5 of model spend a month — then a flat monthly fee by plan size, $49 for up to 10 users to $599 for up to 250 users, with AI credit included. Never per seat, and every feature in every plan.
Pricing

Runbear

Team $99 a month with 20,000 credits and 5 builders; Business $399 a month with 100,000 credits and 10 builders; Enterprise on custom pricing. “Credits, not seats. Start for free.”
10Where it lives

attest_tag

Slack and Microsoft Teams, from one console: channels, threads and DMs in Slack; channel reply chains, group chats and one-to-one chats in Teams. It joins the channels you invite it to, and there is no second place for the team to go.
Threads and streaming

Runbear

“Slack, Discord & Microsoft Teams”, a Team-tier bullet. In Slack it “answers @mentions in any channel it is invited to, and direct messages from individual members”.

“Not stated on their site.” means we read the pages below on 22 Sep 2026 and did not find it — not that the product lacks it. Ask them; a vendor that answers all ten in writing is a good sign about the vendor. If we have something wrong, tell us and we will fix the page.

Pages read: Runbear homepageRunbear pricingRunbear for EnterpriseRunbear docs: SlackRunbear docs: Agent settingsRunbear docs: Billing and creditsRunbear docs: Credits APIRunbear docs: Traces APIRunbear docs: Tool approvalsRunbear docs: Audit logRunbear docs: HIPAA Compliance

Questions

What people ask about Runbear and attest_tag.

Is attest_tag an open-source alternative to Runbear?
Yes. attest_tag is MIT-licensed: one Go binary with the admin console embedded, one database, and a public repository on GitHub. The hosted service runs the same code you can clone and build. Runbear's pages, read on 22 September 2026, do not state a licence or whether the source is available.
Can I run attest_tag on my own servers, and can I run Runbear?
attest_tag self-hosts as one binary: Docker on a laptop or VM, Google Cloud Run, AWS Fargate, Azure Container Apps, or Kubernetes with a Helm chart. Self-hosting has no fee and no limits set by us, and the audit log is then a table in your own database. Runbear's homepage, pricing, enterprise and docs pages do not state a self-hosted, on-premise or private-cloud option.
Which model does each run on?
attest_tag calls any OpenAI-compatible endpoint — OpenRouter, a vendor API, or a model you host — ships on google/gemini-flash, and lets the model be set per workspace, per channel and per thread. Runbear lets the provider be chosen per agent — Claude, GPT or Gemini — and lists “Bring your own LLM key (BYOK)” under its Business tier; its Traces API records “the model” for every run, but whether the model that ran is shown under a reply in Slack is not stated.
What does a turn cost?
attest_tag records every completion's measured tokens and cost and shows them under the reply; !usage shows the month by channel against the budget. A typical answered question on the default model costs about $0.002, a fix job about $0.18. Runbear prices by credits — “A credit reflects the model work a task actually uses” — with 20,000 a month on Team and 100,000 on Business; its docs put a knowledge answer at “10 to 20 credits” and a multi-tool task at “30 to 50 or more”, and a Credits API returns every metered call to the organisation's admins. A per-reply cost shown inside Slack is not stated.
Does Runbear hold a write for a human before it runs?
Their Enterprise page says “Set approval gates on sensitive actions — refunds, deletes, sends” and “The agent drafts, a named human signs off”; the homepage says anyone in the channel can send it or edit first. Whether every write is held by default is not stated. attest_tag holds any method other than GET, HEAD or OPTIONS: it posts what it intends to do with a Confirm card in the thread, expiring after five minutes; a write that grants access goes to an approver's DM instead, and an admin can pre-approve an expected write with an allow rule.
Can someone approve their own request?
Not in attest_tag. A write that grants access leaves the thread and goes to an approver's DM, and nobody approves their own request — enforced three times over, once in SQL. Runbear's pages say a named human signs off, but not who may, or whether a requester can approve their own action; their docs' “Tool approvals” page is admin approval of tool connections, not of individual actions.

More comparisons

The same ten questions, asked of the others.

All comparisons

Try both

Run both in the same channel for a week.

We mean it. attest_tag answers in threads, so it will not collide with whatever else is in the channel, and a week of your own questions settles this faster than any table.

Try it nowDeploy it yourself

Half an hour to set up, one channel to start, and the free plan needs no card. See what it costs.