Multi-account and multi-provider auth in one app
Run every OpenAI account from one desktop app. Sign in once per account, then switch between them instantly, with no re-authentication and no browser loop. A profile can also be bound to an OpenAI-compatible provider instead of an OpenAI account.
Why multi-account matters
- Work vs personal separation: Keep company and personal usage clean.
- Org isolation: Different orgs can have different model access and limits.
- Rate limit planning: Keep heavy jobs on a dedicated account.
- Client workspaces: Use a profile per client to avoid context bleed.
Supported profile types and models
| Profile type | Auth method | Available models |
|---|---|---|
| OpenAI account | OpenAI OAuth | GPT-5.3-Codex, GPT-5.2-Codex, GPT-5.1-Codex-Max, GPT-5.1-Codex-Mini, GPT-5.2 |
| Custom provider | Your provider credential, held in the macOS Keychain | The models you publish on that connection — OpenRouter, DeepSeek, Groq, or a model running on your own Mac |
Setting up your first profile
- Open CodexUse on macOS.
- Click the + button to add a new profile and pick OpenAI account.
- Complete the OpenAI OAuth flow in your browser. CodexUse stores the token locally.
- Name your profile, for example "Work" or "Personal".
Adding more accounts
Repeat the steps above for each OpenAI account. Every profile is independent, so add one per account or org and switch between them at will.
Switching between accounts
From the window
Use the profile cards in the Profiles view or the quick switcher to pick a profile. The switch is quick, no browser needed.
From the system tray
Use the CodexUse tray menu to switch profiles. Each entry shows the profile name so you know which account will run.
Rate limits across accounts
CodexUse shows rate limit usage per profile. When one account approaches its threshold, you can:
- Switch to another profile.
- Enable auto-roll (Pro) to switch automatically.
CLI bundling explained
CodexUse bundles the Codex CLI so you do not need to install it separately. Updates to CodexUse include CLI updates, so you get compatible versions by default.
Common setups
| Use case | Suggested profiles | Why |
|---|---|---|
| Solo developer | Work, Personal | Keep billing and work contexts separate. |
| Consultant | One profile per client | Clean separation across engagements. |
| Team lead | Org profiles by team | Different limits and access per org. |
Related
Do I need API keys to use CodexUse?
Not for an OpenAI profile — that uses the same Codex CLI OAuth login. A custom-provider profile needs a credential for the provider you choose, stored in the macOS Keychain. See the custom model providers docs.
Can I add more than one OpenAI account?
Yes. Each OpenAI profile is tied to one account. Create one profile per account and switch between them from the tray or window.
Which models are available?
On an OpenAI profile: GPT-5.3-Codex, GPT-5.2-Codex, GPT-5.1-Codex-Max, GPT-5.1-Codex-Mini, and GPT-5.2, based on your account access. On a custom-provider profile: only the models you published to it. The two lists never mix.