Codex GPT-6.2 usage limits: Ultrafast uses 6x, credits 3x through February 28
As of November 17, 2026, OpenAI's Codex pricing page puts GPT-6.2 Astra at 6-50 and GPT-6.2 Sol at 17-180 local messages per five hours on Plus. Ultrafast uses 6x of your included usage. Ultrafast credit overages cost 3x through February 28, 2027, then 6x.
SessionWatcher editorial · Published · Updated · Source announcement: 2026-11-17
What the Codex pricing page shows for GPT-6.2
As of November 17, 2026, OpenAI's Codex pricing page lists GPT-6.2 Astra and GPT-6.2 Sol on Plus, Pro, Business, Enterprise and Edu. You can use them in ChatGPT Work on web, mobile and desktop, and in the Codex app and CLI. Both models are off by default in Enterprise and Edu workspaces until an administrator turns them on.
The page estimates local messages per five-hour period. On both Plus and Standard Business, GPT-6.2 Astra gets 6-50 and GPT-6.2 Sol gets 17-180. For comparison, GPT-6.1 Sol gets 15-160 and GPT-6 Astra gets 5-45. Pro plans currently have no five-hour limit, and weekly limits may also apply. OpenAI says these estimates are not fixed message limits.
Sources: Pricing | ChatGPT Learn
Fast and Ultrafast rates for GPT-6.2
Fast mode on GPT-6.2 Astra and GPT-6.2 Sol uses 2x the Standard rate, from both included usage and credits or pay-as-you-go. Ultrafast uses 6x of included usage. Ultrafast overages paid with purchased credits, and eligible Enterprise pay-as-you-go usage, cost 3x the Standard rate through February 28, 2027. Both rise to 6x starting March 1, 2027, Pacific time. The page lists Ultrafast for eligible Pro, Enterprise and Edu plans, with access on Pro $500.
Sources: Pricing | ChatGPT Learn
What it means for your usage and bill
Your plan's included usage is used first, so an Ultrafast turn on GPT-6.2 drains it six times as fast as Standard. Once you reach your included limits, you keep working on credits. Until February 28, 2027, Ultrafast on credits costs half its included-usage rate. From March 1, the same work costs twice as many credits.
At Standard speed, GPT-6.2 Sol costs 50 credits per million input tokens, 2.5 per million cached input tokens and 250 per million output tokens. GPT-6.2 Astra costs 200, 10 and 1,000 credits. With USD usage-based billing, Astra costs $8, $0.40 and $40 per million tokens and Sol costs $2, $0.10 and $10.
Sources: Pricing | ChatGPT Learn
How to check your limits
OpenAI points to the usage dashboard for current limits and reset times. In an active Codex CLI session, /status shows your remaining limits.
Sources: Pricing | ChatGPT Learn
AI assisted reporting, checked against the linked official sources. Source pages checked 2026-11-17. Editorial process and corrections.