Skip to main content
OpenWork Models is a set of models that OpenWork picks and runs for you. One subscription covers your whole organization. Nobody adds a provider or pastes an API key: members pick a model from the OpenWork group in the app and start working. It’s the quickest way to give a team AI models. If your company already pays for OpenAI, Anthropic, Google or another provider, you can add your own providers instead, or use both.

​
Where to find it

In OpenWork Cloud, open AI Gateway and select the OpenWork Models tab. Seeing this tab needs the View OpenWork Models settings permission. Organization owners and admins have it by default.

​
Turn it on

1

Subscribe

On the OpenWork Models tab, click Subscribe. You go to checkout to pay, then come back to OpenWork. Subscribing needs the Manage billing permission.
2

Enable (if asked)

If the tab shows Enable after you subscribe, click it to turn the models on for your organization. This needs the Manage OpenWork Models permission.
3

Pick a model

Members choose a model from the OpenWork group in the desktop app and start a task. There is nothing for them to install or set up.
Once it’s on, the button changes to Manage subscription, which opens Settings › Billing.
If the tab says OpenWork Models isn’t sold where you are, there’s no Subscribe button. Click Add a team provider to share your own provider accounts instead.

​
What it costs

The tab shows the price next to the button: $10 / user / month, billed per active member. See Pricing and licensing for how it fits with your plan. Model usage is included up to your shared usage limits.

​
Included usage

When OpenWork Models is on, the tab shows Shared usage limits. Usage is shared across all active members of your organization, so anyone’s work draws from the same allowance. Adding active members raises it. There can be up to three limits, each with its own reset time:
  • 5 hour usage limit: resets during the day.
  • Weekly usage limit: resets on the date shown.
  • Monthly usage limit: resets on the date shown.
Each card shows how much is left and when it resets.

​
Which models are included

The Models list on the tab shows every model in the lineup, what each one is best for (for example research, long documents, or spreadsheets), and its model ID. Every member can use all of them. OpenWork updates the lineup over time, so the tab is always the current list.

​
Who can use it

Everyone in your organization. OpenWork Models isn’t granted per team or per person the way your own providers are. To keep members on approved models only, open AI Gateway › AI Providers, click Change next to Who can use models, and choose Only models you provide. Members then see only what you set up in AI Gateway and OpenWork Models, and can’t add their own provider keys.

​
See how much is used

  • The Shared usage limits cards on the OpenWork Models tab show what’s left of the included usage.
  • The Overview tab shows spend and tokens across all Gateway providers, including OpenWork Models, once you have at least one provider of your own.

​
FAQ

No. They’re separate providers with separate subscriptions. See OpenWork vs OpenCode models.
Yes. Members see both in the app. Your own providers follow the access rules and limits you set in AI Gateway.
OpenWork Models isn’t offered everywhere. Organizations with a signed data processing agreement can’t use company-provided OpenWork Models, and some self-hosted deployments don’t sell it. You can still add your own providers.