Fetching from the wire…
01
02
03
04
05
06
07
08
Models2026-09-25 · source-backed
RuntimeWire, citing TestingCatalog, reported on September 24 that a Pro Max plan would add faster Work and Codex and possibly higher limits, with speculation about Cerebras hardware. OpenAI hasn't confirmed it and DevDay is September 29. (RuntimeWire) Codex PR #47971, merged September 25, recognizes a promax plan in auth, account responses and rate limits, relabeling tiers as Pro (prolite), Pro (More) (pro) and Pro (Max) (promax). PR #47932 removed gpt-5.4 from the bundled and Bedrock catalogs including GovCloud, prompting saved selections to migrate to GPT-6 Sol. The client code isn't confirmation of pricing, but it does confirm the tier exists as a concept inside OpenAI.
Each link below shares sources, entities, or timing with this story.
codenotch is a Swift menu-bar app at 1,011 stars, created September 5, that reads Claude Code's /usage output (falling back to the OAuth token in the login keychain), Cursor's local SQLite session state, Codex's ChatGPT usage endpoint for 5-hour and weekly limits, plus Antigra...
First-party pricing, counts toward AWS commitments, Codex via CLI and IDE plugins for VS Code, JetBrains, and Xcode, across commercial and GovCloud (AWS). This removes the procurement and compliance wall for AWS shops that couldn't touch OpenAI under existing contracts. Distri...
It's the customer-run half of OpenAI's Secure MCP Tunnel, letting a private or localhost MCP server reach ChatGPT, Codex, the Responses API and Agent Builder without exposing it publicly. 422 stars, v0.0.14 tagged September 1, pushed this morning. GitHub Pre-1.0 numbering and...
b-nnett/codex-subscription-router, created August 16 and already past 200 stars, builds a locally patched copy of the official ChatGPT desktop app and fans its single app-server connection out to one official Codex child process per account through a Go multiplexer, each with...
Three things happened this month that only make sense together. Agent Plugins 1.0 shipped co-signed by six competitors: AWS, Anysphere, Microsoft, OpenAI, Vercel and Google (GitHub Changelog). It makes skills-plus-MCP bundles portable across clients. OpenAI's August 11 Codex c...
A merged PR is not a changelog. That's the lesson. openai/codex#33972 landed July 19 and caps the effective GPT-5.6 window in Codex at 272,000 input tokens plus 128K reserved output. It was 372K. The change shipped as model metadata. No blog post, no deprecation notice, no mig...
MindPattern daily
One email a day at 7 AM. Sources and a take on every story. Unsubscribe anytime.