GPT-6.1 Sol Missing in ChatGPT? Check Work and Codex

GPT-6.1 Sol missing or not supported with a ChatGPT account? Check the Codex client, login method, workspace access, and a small recovery test.

If GPT-6.1 Sol is missing from the normal ChatGPT Chat model picker, that alone does not establish a rollout failure. OpenAI's current model documentation places GPT-6.1 Sol, GPT-6 Sol, and GPT-6 Luna in Work and Codex, not Chat. Start by checking which surface you have open.

If Codex instead says the model is not supported with a ChatGPT account, a missing picker is not the whole problem: the request was rejected on that access path. Typing the model ID again does not grant permission. Use the diagnostic steps below before buying another plan or changing to API billing.

Availability and troubleshooting reviewed October 14, 2026. This is a source-based diagnostic guide, not a tested repair for every account. A model announcement, a paid plan, and a selectable model are three different pieces of evidence.

Which model was released when?

The API changelog records GPT-6 Sol and Luna on September 22 and GPT-6.1 Sol on September 29. Do not treat access to last week's Sol as proof that the new 6.1 version is enabled. Likewise, an article written before September 29 may describe GPT-6 Sol accurately without covering the new rollout.

Use the GPT-6.1 Sol pricing and upgrade guide for API rates and protocol changes. This page focuses on locating and verifying the model in a working client.

Check the surface, account, and workspace in that order

What you observeWhat to check next
You are in an ordinary Chat conversationOpen the appropriate Work or Codex surface; a Chat picker is not the availability test for these models.
The model is absent in Work or CodexCheck the signed-in account, plan, client version, and rollout conditions.
A personal account has access but a team workspace does notCheck the selected workspace and administrator model policy.
The CLI rejects the IDVerify the exact ID, client support, active authentication, and permitted provider route.
An API request succeeds but the subscription picker is emptyInspect those access paths separately; an API project does not prove subscription access.

The launch rules are version-specific. OpenAI documents GPT-6.1 Sol for Plus, Pro, Business, Enterprise, and Edu; Free and Go are excluded at launch. Enterprise and Edu administrators must enable it. Luna's earlier rollout has different eligibility. These conditions may change, so use the live model availability page when diagnosing access.

Do not buy a higher plan before establishing which check failed. A client that lacks the new catalog, a workspace policy, and a plan restriction require different fixes.

Before continuing, choose the task you actually need. If you want GPT-6.1 Sol in OpenAI's own ChatGPT Work or Codex, follow the official checks below. If you want a coding Agent to work on project files in a shared environment, check Codex's model selection and workflow on Macshop first. Its available combinations are determined by the live Workspace selector; Macshop does not unlock a missing model in your native ChatGPT account.

“Not supported when using Codex with a ChatGPT account”: what to check

A reported error is:

text
The 'gpt-6.1-sol' model is not supported when using Codex with a ChatGPT account.

An issue in OpenAI's Codex repository records this message. It is a user report, not an official finding that all ChatGPT subscriptions lack access. Another cross-client report describes different model availability under the same account; it does not establish a universal update or cache-clearing fix.

Use this sequence to isolate the failure:

  1. Record the failing client. Is it the desktop app, CLI, IDE extension, or Work on the web? For the CLI, record codex --version. Updating a separately installed CLI is not evidence that the desktop app's bundled client changed; use the affected client's official update path.
  2. Verify the identity actually in use. In the CLI, codex login status reports the authentication method. In the app or web interface, inspect the account and selected workspace. OpenAI's authentication guide distinguishes ChatGPT sign-in from API-key billing. If the identity is wrong, use the documented sign-in flow for the intended account; do not delete credential files or work around managed restrictions.
  3. Check eligibility and workspace controls. Compare that identity with the documented model availability above. If an administrator must enable access, ask them to verify it. A successful API request under another project does not answer this question.
  4. Make one controlled comparison. With the same account and workspace, check another supported client you already use. If it works there but not in the original client, record that contrast for support. It narrows the investigation; it does not prove a specific root cause.
  5. Confirm recovery, not just selection. Select the authorized model and try a small, non-sensitive request that needs no file changes, such as “Reply with OK.” A completed response without the rejection confirms that request succeeded. A picker entry alone does not. Record any model-switch notice; success after switching models does not confirm Sol access.

If it still fails, send support the exact error, time and timezone, affected client version, account/workspace type, and comparison result. Redact keys, tokens, private files, and account identifiers in public reports. Do not repeatedly retry the full coding task.

For a different message containing not enabled, keep the complete wording rather than assuming it is the same failure. We have not verified a universal repair for that message. To continue urgent work, use a model already available to the authorized account and review the result; this does not restore the rejected model.

The Codex workflow guide covers the separate question of choosing a coding workspace.

Check Ultrafast eligibility separately

OpenAI's October 5–9 update adds GPT-6.1 Sol Ultrafast in Work and Codex for Pro $500 and eligible Enterprise and Edu plans. Enterprise owners must enable it; supported inference residency includes the United States and Europe (EEA plus Switzerland). Access to ordinary Sol does not establish access to this mode. Check current pricing before enabling a faster setting.

Check saved GPT-5.5 selections around retirement

The official changelog, rechecked October 14, schedules GPT-5.5 retirement from ChatGPT, Work, and Codex for October 14, 2026, across plans. That notice excludes the OpenAI API. The published date does not establish when an individual account's picker changed. If an older saved selection disappears, check this scope before treating it as an account failure.

For a Codex workflow using ChatGPT sign-in:

  1. Confirm that this is the subscription-backed workflow, not a separately billed API integration.
  2. Review workspace defaults, saved selections, managed configuration, custom agents, schedules, and scripts that still request gpt-5.5. Choose a replacement actually available to that account and client; changing one visible picker does not update every saved task.
  3. Run a small representative task and check its selected model, output, and tool behavior before relying on the next scheduled run. Keep a record of the configuration change, without credentials.

Do not change an API production integration solely because of this ChatGPT retirement notice. Its model lifecycle must be checked separately. Nor should you assume that selecting a newer model preserves the old model's cost, limits, or behavior.

Select the exact model in Codex

In an eligible, supported CLI, you can start with:

sh
codex --model gpt-6.1-sol

Or choose a model through /model. OpenAI's CLI model documentation describes model selection. Confirm the selected model before evaluating the result, especially if you have saved defaults or a separately configured provider.

A selection is useful evidence about the requested model. It is not proof that every child task or provider response used the same ID. If that distinction matters, preserve the available request/response metadata and task notices, without publishing credentials. The model-routing verification guide explains the limits of each signal.

A subscription and the API answer different questions

When comparing access, write down the client, account, workspace, model ID, and who pays. This five-field note prevents an API test from being mistaken for a subscription test or a managed workspace from being mistaken for a native OpenAI account.

For example, a developer might use a ChatGPT plan in the desktop app and an API project in a server integration. The same person owns both, but the permissions, limits, model availability, and bills remain separate. See the Codex pricing guide before changing the payment route.

Pick a model for the task after access works

Our suggested starting point is a small task with a visible finish line: repair one reproduced bug, explain one code path, or update one document against supplied sources. Save the starting files and define what the output must pass before running it.

Record whether the result passed, how much correction it needed, and which usage meter changed. This produces a useful first comparison without pretending one run establishes a universal winner. For ongoing selection, use the model evaluation checklist.

Next step: verify your setup with one real task

First decide where the task needs to happen. For an OpenAI-native Work feature, use OpenAI's own access checks. Macshop is a separate product and does not grant a native ChatGPT subscription.

For a coding project, a Workspace is a work area that keeps your project files and lets an Agent act on them. Confirm that your intended Agent and model are compatible in the current selector, then check the relevant model costs and allowance. A model announcement alone does not establish its availability or cost in Macshop.

Once those conditions fit, create a workspace and try your first task. Bring one file you are allowed to share and give a read-only request: “Explain what this file does, identify one possible improvement, and suggest how to verify it. Do not edit files or install anything.”

Check that the Agent actually read the intended file and that its explanation matches it. Then decide whether to approve an edit. Creating an account or selecting a model is not evidence that the task ran; this example is a suggested first test, not a measured customer result or a promise of free usage.

If Codex is not the right Agent for your task, consult the supported Agent directory. If you are still comparing where the project should run, use the Workspace selection guide.