C/ YOUR PRIVATE RUNTIME

A place for
your Codex.

One token. Your models. Settings that stay where you put them.

Service token

The same token authorizes API calls. Your Codex account credentials stay on the server.

PRIVATE
BY DESIGN.

CODEX / CONTROL ROOM

C/ Private Codex

CONTROL ROOM

Connecting…

One runtime. One key.
Ready for your next idea.

PERSONAL DEVELOPMENT INFRASTRUCTURECodex CLI

SET ONCE. CALL ANYWHERE.

Your profiles.

Keep model settings here. Send the profile name from your code.

SAVED PROFILES

PROFILE SETTINGS

Default

DEFAULT

Profile name

Model

Reasoning effortHow much reasoning the model applies.

Service tierSpeed preference, independent of reasoning.

Instructions optional

Advanced settings

Response length

Reasoning summary

Timeout (seconds)

Output JSON schema optionalLeave empty for text. Structured output is validated against this schema.

USE IT FROM YOUR APP

Send a prompt. We apply the profile.

TRY BEFORE YOU INTEGRATE

Playground.

A real call to your Codex runtime, using your saved settings.

ProfileYour prompt

No requests running.

RESPONSE

Your response will appear here.

Applied settings & usage

THE ENGINE ROOM

Service.

Your runtime, its connection, and the key that opens it.

RUNTIME

Checking…

Model calls run in isolated, ephemeral threads. Tools and server file access are disabled.

API ACCESS

One service token.

Send it as a Bearer token. Every valid token holder can use all profiles and administer this service.

Token fingerprint

Rotating the token immediately signs out all sessions and invalidates the previous key.

INTEGRATION

Bring your existing SDK.

Native API for profiles and documents. OpenAI-compatible endpoints for text, JSON and streaming.

Native base URL
OpenAI SDK base URL
Profile as model: profile:your-profile-name

Hosted tools, server-side conversation storage and unsupported parameters return an explicit error.

PRIVATE CODEX Built around your workflow.

Your replacement token

Save this now. The old token no longer works.