Get an API key from OpenAI

Make the key on the OpenAI Platform, put a hard limit on the month's spend, then paste it into Runner. Every step at OpenAI comes from OpenAI's own documentation.

Updated

The short version

  • Make a standard API key on the OpenAI Platform's API keys page, with an expiration date, as OpenAI recommends.
  • Add credits on Billing, then turn on a hard spend limit on Limits, so the month stops where you chose.
  • In Runner, an owner or admin pastes it under Settings, then LLM providers, presses Test, then Connect.
  • Rather skip all of it? With AI included there is nothing to set up, and you can change your choice later.

Sign in to the OpenAI Platform, and add credits

Keys for OpenAI's models live on the OpenAI Platform, OpenAI's dashboard for its API, at platform.openai.com. Your organization pays for what its keys use from credits it buys there in advance.

  1. Sign in

    Go to platform.openai.com and sign in to the dashboard.

  2. Add credits

    Open Billing, at platform.openai.com/settings/organization/billing, and add credits.

When the credits run out

OpenAI refuses requests once the organization's prepaid credits are used up, and they resume when you add more on Billing.

Create a key, with an end date

The kind of key OpenAI makes for an application like Runner is a standard API key, made on the API keys page.

  1. Open API keys

    Open API keys, at platform.openai.com/settings/organization/api-keys.

  2. Create the key, with an expiration date

    Create a new key. OpenAI strongly recommends setting an expiration date when you do, so the key stops on a date you chose.

  3. Keep a copy somewhere safe

    Copy the key and store it in a secure location, such as a password manager, never in a document or a message.

Before the date comes

Make a replacement before the key expires, move Runner to it, then revoke the old one. The last section below shows how.

Set a hard spend limit at OpenAI

A spend limit caps your organization's spend for the month. Making it a hard limit is what stops requests when spend reaches it.

  1. Open Limits

    Open your organization's Limits page, at platform.openai.com/settings/organization/limits.

  2. Edit the spend limit

    Under Spend, select Edit spend limit and enter a Monthly spend limit.

  3. Make it hard, and save

    Turn on Enforce a hard limit, so responses stop at the limit, then select Save.

Alerts, and how exact the limit is

Spend alerts on the same page tell you when spend passes an amount, but an alert stops nothing; the hard limit does. OpenAI applies it as spend is recorded, so the month can end slightly over. It resets with the next monthly cycle, or sooner if you raise or remove it.

Paste the key into Runner, under Settings, then LLM providers

The key belongs to your Runner organization, so every product in it can use the key, and only an owner or admin can connect one. Runner encrypts it before storing it and never shows it again.

A key you connect is used before credits, even on AI included.

  1. Open LLM providers

    In Runner, open Settings, then LLM providers, under What the AI uses.

  2. Find the OpenAI card

    If it's folded under More providers, open it. Paste the key into OpenAI API key.

  3. Press Test, then Connect

    Test asks OpenAI whether it accepts the key, without using any credit. When it says OpenAI accepted it, press Connect.

On Runner's LLM providers page, paste the key into the OpenAI API key box, press Test, then Connect.

Keep the key secret

In OpenAI's words, your API key is a secret: don't share it with others. Anyone else who needs access gets a key of their own.

  • Keep it out of code, files and public repositories, and store it in a secure location instead.
  • Never put it in anything a browser or an app downloads.
  • Revoking a key takes effect within a few seconds, so a key you think has leaked can be stopped at once.
  • Paste it only into Runner's key box, never into a chat, an email or a support ticket.

Swap the key for a new one, or switch it off

OpenAI's advice is a regular rotation: create a replacement, move to it, and revoke the old key once the replacement works.

  1. Make the replacement

    On API keys, create a new key with its own expiration date.

  2. Swap it in Runner

    On the OpenAI card in LLM providers, press Disconnect, then Confirm disconnect, and paste the new key, Test it and Connect. Disconnecting stops your team using the old key at once.

  3. Revoke the old key

    Once Runner is connected with the new key, revoke the old one on API keys. OpenAI says a revocation takes effect within a few seconds.

Rather not make a key at all?

With AI included, your team runs on the AI credits your plan includes, with nothing to set up: no API keys, no vendor accounts, no billing with an AI company.

Every plan comes two ways, AI included or your own key, and you can change your choice later in Settings.

AI included or your own key: help me choose

Give your business a growth team.

Create your account and pick a plan. Your team is built once the plan is paid.

  • AI models and keys

    Get an API key from Anthropic (Claude)

    Claude API key from Anthropic, step by step: make a Claude Console account, buy credits, create the key, set a spend limit, and connect it to Runner.

    Updated

  • AI models and keys

    Get an API key from Google (Gemini)

    Gemini API key from Google, step by step: create the key in Google AI Studio, set up billing, cap the month's spend, and connect it to Runner.

    Updated

  • AI models and keys

    Get an API key from xAI (Grok)

    Grok API key from xAI, step by step: buy credits in the xAI console, create the key, keep spending inside what you bought, and connect it to Runner.

    Updated

Trademarks of their owners. No affiliation or endorsement.