For the complete documentation index, see llms.txt. Markdown versions of documentation pages are available by appending .md to the page URL.
Primary navigation
Theme

Fast and Ultrafast modes

Ultrafast

In ChatGPT Work and Codex, Fast and Ultrafast generate tokens faster than Standard on the same model. Both use more of your plan’s allowance or credits.

Choose a speed in the composer

In ChatGPT Work on the web or in the Codex desktop app, open the model or effort picker at the bottom of the composer.

  • If Ultrafast is available to you, select the lightning bolt to open the speed menu, then choose a speed.
  • If your only options are Standard and Fast, selecting the lightning bolt switches between them immediately.

In the advanced picker, select Speed. In ChatGPT Work on mobile, open the model picker and use the speed setting.

Try switching speeds in the desktop picker below.

Speed

Fast mode

Fast mode supports GPT-6.2 Astra and GPT-6.2 Sol. You can also use it with GPT-6.1 Sol, GPT-6.1 Terra, GPT-6 Astra, GPT-6 Sol, and GPT-6 Luna. On GPT-6.2 Astra and GPT-5.6, Fast generates tokens at about 1.5 times the speed of Standard.

With ChatGPT sign-in, you can also use Fast in the Codex CLI and IDE extension. See Fast pricing for the usage rate for your model.

Ultrafast mode

On GPT-6.2 Astra and GPT-6.2 Sol, Ultrafast generates tokens at up to 4.5 times the speed of Standard on the same model. Choose Ultrafast in the composer in ChatGPT Work or the Codex app, or use the Codex CLI.

ChatGPT Work and Codex share pricing, credits, and usage limits. See Ultrafast pricing for included usage, purchased credits, and eligible pay-as-you-go rates.

GPT-6 Astra and GPT-6.1 Sol also support Ultrafast on Pro $500 and eligible Enterprise and Edu plans. In Codex, GPT-6 Astra generates tokens at up to eight times its Standard speed with Ultrafast.

Ultrafast plan and workspace requirements

Enterprise workspaces need a credit-based or USD usage-based agreement. Edu workspaces need a credit-based plan.

Workspace owners must enable the GPT-6.2 models they want members to use. Ultrafast is a separate permission and is off by default. Owners can enable it for selected users or the whole workspace in workspace permissions. Per-user spend controls apply.

For GPT-6.2 Astra, GPT-6.2 Sol, and GPT-6 Astra, Ultrafast supports inference residency only in the United States. For GPT-6.1 Sol, it also supports Europe (EEA and Switzerland).

Codex CLI controls

In the CLI, type / to see which commands are available for your model. Use /fast or /ultrafast to choose a speed. Repeat the active command to return to Standard. The CLI saves your selection. To show Fast mode in the footer, use /statusline.

API billing

When you sign in to Codex with an API key, you pay API token rates. ChatGPT credit multipliers apply only when you sign in with ChatGPT. See Ultrafast mode in the API for supported models, request configuration, and pricing.

Retirement and migration

GPT-5.5 retired from Codex for users signed in with ChatGPT on October 14, 2026, at noon PDT (19:00 UTC).