The short version
- December 2026 Overhaul: ChatGPT and Codex Merge Into One Shared Token Pool (Effective Date: Starting December 2026; Eligible Plans: ChatGPT Plus, Pro, and Business tiers).
- Ending Split Quotas for Heavy Codex Developers and Pure Chat Users Alike (Metering Change: Replaces separate caps with one shared token pool).
- Managing Context Size and Reasoning Effort Under Token-Based Subscription Metering (Key Benefit: Allocate up to 100% of monthly quota to either Codex or Chat).
At a glance
| Effective Date | Starting December 2026 |
|---|---|
| Eligible Plans | ChatGPT Plus, Pro, and Business tiers |
| Metering Change | Replaces separate caps with one shared token pool |
| Key Benefit | Allocate up to 100% of monthly quota to either Codex or Chat |
| Category | Dev & Open Source |
| Creator / publisher | Veee |
| Official link | openai.com ↗ |
| Heat | 🔥 4.6K Heat |
| Announced | |
| Data updated |
December 2026 Overhaul: ChatGPT and Codex Merge Into One Shared Token Pool
OpenAI announced that starting in December 2026, paid ChatGPT Plus, Pro, and Business subscriptions will replace separate message and coding caps with a single unified, token-based usage pool shared across ChatGPT and Codex.
Ending Split Quotas for Heavy Codex Developers and Pure Chat Users Alike
Under the unified meter, developers who spend most of their time in Codex can allocate nearly their entire monthly subscription allowance to repository tasks, while general users can spend 100% of their pool inside ChatGPT conversations.
Managing Context Size and Reasoning Effort Under Token-Based Subscription Metering
Because multi-file repo scans and high-reasoning turns consume significantly more tokens than short chat messages, compacting long sessions and scoping file reads will help stretch the unified monthly pool much further.
FAQ
What are the verified facts and timeline behind ChatGPT & Codex Unified Token Pool?
Effective Date: Starting December 2026; Eligible Plans: ChatGPT Plus, Pro, and Business tiers; Metering Change: Replaces separate caps with one shared token pool. Under the unified meter, developers who spend most of their time in Codex can allocate nearly their entire monthly subscription allowance to repository tasks, while general users can spend 100% of their pool inside ChatGPT conversations.
What are the key implications and what happens next?
Key Benefit: Allocate up to 100% of monthly quota to either Codex or Chat. Because multi-file repo scans and high-reasoning turns consume significantly more tokens than short chat messages, compacting long sessions and scoping file reads will help stretch the unified monthly pool much further.
