GPT-6.2 Sol is disabled by default in Enterprise and Edu workspaces. Administrators must enable it in workspace model settings, and members must also have the appropriate access to ChatGPT Work or Codex.
Suitable tasks: Make GPT-6.2 Sol available to members of an Enterprise or Edu workspace and verify availability in each client.
Unsuitable tasks: Configure model access for personal plans, OpenAI API organizations, or projects. Those are not controlled by ChatGPT workspace settings.
Applicable model version: GPT-6.2 Sol.
Applicable clients, agents, or APIs: ChatGPT Work and Codex (verify each client actually in use separately). API key sign-in follows the permissions of the associated API organization and project.
Recommended reasoning levels and parameters: The source does not provide specific recommended reasoning levels or parameters for GPT-6.2 Sol.
The following administrator checklist is based on the official guidance:
Enable GPT-6.2 Sol under Admin Console > Workspace settings > Models.
Confirm that members have access to ChatGPT Work or Codex. Enabling the model alone does not grant access to those products.
If Ultrafast is needed, check whether the workspace is eligible and enable the option separately.
After saving, reopen the model access settings and verify the saved selection.
Ask affected members to confirm that the model appears in the supported clients they use.
Follow this sequence: “Enable the workspace model → check member product access → verify the saved state → validate each client.” If a member still cannot see the model, first check their sign-in identity, workspace role, and client support.
The official page says GPT-6.2 Astra and GPT-6.2 Sol are both disabled by default when launched in Enterprise and Edu, and administrators can enable them in workspace model settings.
Members also need access to ChatGPT Work or Codex to select the model there.
Eligible workspaces can enable Ultrafast separately.
ChatGPT workspace settings do not automatically apply to the Codex client, Codex cloud, or the OpenAI API. Model access for API key sign-ins is determined by the associated API organization and project.
The source page does not show an author or publication date, and provides no specific recommended reasoning level for GPT-6.2 Sol.
This workflow is for Enterprise and Edu workspace administrators. Personal plans do not use the workspace enablement process described here. Model availability for a specific personal plan should be confirmed based on the relevant product and account permissions. Enabling a model in a workspace does not mean members have access to Work or Codex, nor does it grant API access.
In an Enterprise or Edu workspace with the required administrative permissions, save the settings using the checklist above, then ask affected members to check the clients they use. Model availability and permissions may vary by workspace, role, and client.
GPT-6.2 Sol