# AI Updates: ChatGPT and Codex Share One Limit, a Doubtful Mythos Leak, and Unslop OpenAI plans to put ChatGPT and Codex on one shared token allowance for Plus, Pro and Business from December, but has not published the numbers, so nobody… **Published:** 2026-11-02 · **Author:** Yash Thakker **Source:** https://explainx.ai/newsletter/ai-updates-chatgpt-and-codex-share-one-limit-a-doubtful --- > **In short:** OpenAI plans to put ChatGPT and Codex on one shared token allowance for Plus, Pro and Business from December, but has not published the numbers, so nobody can yet tell whether heavy users will gain or lose. ## One Pool for Paid Plans ### ChatGPT and Codex Usage Limits Merge in December: One Token Allowance OpenAI says that from December, Plus, Pro and Business plans will share one usage allowance across ChatGPT and Codex. The allowance is measured in tokens, which are small pieces of text. Instant chat will not count against it. OpenAI has not published exact limits or a date, so heavy users cannot yet compare. **What you can do:** Note how often you hit a reset today. Then open Settings, then Usage, when the change arrives, and compare. [Source](https://explainx.ai/blog/chatgpt-codex-usage-limits-merge-shared-token-allowance-december-2026) ## AI News Briefs ### The "Leaked Anthropic Mythos Checkpoint" on Hugging Face Looks Fake: What Is Claimed vs. Verified A Hugging Face page says it holds a leaked early copy of Anthropic's next Mythos-class model. A public audit found 43 pieces of it that are random numbers, rebuilt exactly from a seed. Anthropic has not commented. Treat it as strongly doubted, and do not download or run it. [Source](https://explainx.ai/blog/alleged-anthropic-mythos-checkpoint-numbat-hugging-face-fake-random-weights-2026) ### unslop: The Claude Skill That Fixes Vibecoded, AI-Looking UI unslop is a free Claude skill, a reusable set of instructions, that tidies up screens that look AI-made. It looks for purple gradients and emoji labels, proposes a direction, then fixes spacing, color and copy in nine ordered passes. It is days old, so try it on a copy. [Source](https://explainx.ai/blog/unslop-claude-skill-fix-vibecoded-ui-heydamir-github-trending-guide-2026) ### Musk Says Grok Bot Cost "Unsustainably High" as Weekly Limits Bite: What Is Verified Startup Fortune reports that Elon Musk said on X the cost of serving Grok Bot was "unsustainably high relative to revenue." He gave no figures, limits or dates. Users report using up weekly allowances in days, but xAI does not publish the allowance, so no cut is confirmed. [Source](https://explainx.ai/blog/musk-grok-bot-serving-cost-unsustainably-high-weekly-usage-limits-what-is-verified-2026) ## One Useful Idea ### Claude Usage Limits in 2026: Every Change Explained (Timeline) Claude has two separate limits: a rolling five-hour window for short bursts, and a weekly cap for sustained use. Anthropic has changed these numbers several times since August 2025, so remembered figures go stale. This guide dates each change and lists habits that stretch any plan, such as batching related questions into one message. **What you can do:** Check your own usage panel before choosing a plan, and ask related questions in one message instead of several. [Source](https://explainx.ai/blog/claude-usage-limits-2026-timeline-explained) ## Upcoming Workshops ### Grok Bot & 24/7 AI Agents (Duet, Dots, Muse) A free 90-minute live session for beginners on Grok Bot, xAI's always-on agent app. It covers a live demo, safety rules for agents that run around the clock, and a comparison with Claude Duet, ChatGPT dots and Meta Muse. No coding needed. [Source](https://explainx.ai/workshops/grok-bot) ### Claude for Work: Become Your Team’s AI Person ![Claude for Work: Become Your Team’s AI Person](https://explainx.ai/workshops/covers/square/claude-for-work.webp) *Live workshop · Nov 28–29, 2026* Two live evenings for professionals who want to use Claude well at work. You build a knowledge base, an agent skill, a prompt library and your own Artifacts, and connect Claude to the apps you use. No coding or technical background is needed. [Source](https://explainx.ai/workshops/claude-for-work) ## Reader poll Would one shared usage limit for chat and coding help you? - Yes, it is simpler - No, I prefer separate limits - It depends on the numbers - Not sure ## Sources - [ChatGPT and Codex Usage Limits Merge in December: One Token Allowance](https://explainx.ai/blog/chatgpt-codex-usage-limits-merge-shared-token-allowance-december-2026) - [The "Leaked Anthropic Mythos Checkpoint" on Hugging Face Looks Fake: What Is Claimed vs. Verified](https://explainx.ai/blog/alleged-anthropic-mythos-checkpoint-numbat-hugging-face-fake-random-weights-2026) - [unslop: The Claude Skill That Fixes Vibecoded, AI-Looking UI](https://explainx.ai/blog/unslop-claude-skill-fix-vibecoded-ui-heydamir-github-trending-guide-2026) - [Musk Says Grok Bot Cost "Unsustainably High" as Weekly Limits Bite: What Is Verified](https://explainx.ai/blog/musk-grok-bot-serving-cost-unsustainably-high-weekly-usage-limits-what-is-verified-2026) - [Claude Usage Limits in 2026: Every Change Explained (Timeline)](https://explainx.ai/blog/claude-usage-limits-2026-timeline-explained) - [Grok Bot & 24/7 AI Agents (Duet, Dots, Muse) — Free Live Session](https://explainx.ai/workshops/grok-bot) - [Claude for Work](https://explainx.ai/workshops/claude-for-work) --- From [explainx.ai](https://explainx.ai/newsletter). Read online, comment and subscribe: https://explainx.ai/newsletter/ai-updates-chatgpt-and-codex-share-one-limit-a-doubtful