Fetching from the wire…
01
02
03
04
05
06
07
08
Public story · 2026-07-19 · high
The release notes don't say how long the window was cut or how many users it hit before OpenAI fixed it on July 18.
Why now: OpenAI's July 18 release notes are the only public account of the fix as of July 19.
OpenAI restored full 272K-token context windows to GPT-5.6 Sol, Terra, and Luna in Codex on July 18, per its release notes.
If you benchmarked those three models in Codex over the past few weeks, you may have graded a bug, not the model. The release notes bundle the fix with other changes and don't say how long the window had been running short.
A 272K context window is a specific, testable number. When it quietly shrinks, everything measured against it shifts too. Any claim about how well Sol, Terra, or Luna handle long documents or big repos was measuring degraded infrastructure, not the model's real ceiling. OpenAI's notes don't say what caused the shrink or whether it hit every Codex user the same way.
Without a disclosed start date, there's no way to know which past runs on Sol, Terra, or Luna hit the truncated window. The safest move is rerunning any long-context eval from before July 18.
Each link below shares sources, entities, or timing with this story.
GPT-5.6 Luna went to $0.20 input / $1.20 output per million tokens on July 30. That's an 80% cut. Terra dropped 20%. Luna's input now undercuts Gemini 3.1 Flash-Lite ($0.25/$1.50) and sits at one-fifth of Claude Haiku 4.5's $1 input. Simon Willison covered the announcement and...
Four frontier models. Five sealed engineering problems. The result everybody will quote is that Claude Fable 5 won. The result that should actually change how you work is buried three-quarters down the page. JuliaHub published an evaluation on July 30 running four frontier mod...
OpenAI launched the GPT-5.6 family on July 14: Sol (flagship), Terra (cost-optimized), and Luna (fast tier), live across ChatGPT, Codex, and the API the same day after a US-government-requested delay for security review. The numbers are loud. Sol scored 53.6 on Agents' Last Ex...
On July 9 the standalone Codex app merged into ChatGPT desktop for macOS and Windows, keeping a dedicated coding surface alongside Chat and Work modes, plus inline diff editing, side-panel PR review, faster Computer Use on GPT-5.6, and multi-repository projects. The tiering: S...
July 9, across VS Code, Visual Studio, Copilot CLI, the cloud agent, github.com, GitHub Mobile, JetBrains, Xcode, and Eclipse. Sol is the high-reasoning tier at $5/1M in, $30/1M out, gated to Pro+/Max/Business/Enterprise. Terra is the balanced default at $2.50/$15. Luna is fas...
Thibault Sottiaux at OpenAI published an investigation into "a handful of reports where GPT-5.6 unexpectedly deleted files," finding it happens most commonly when full access mode is enabled in Codex. Simon Willison relayed it. A frontier lab publishing a first-party post-mort...
MindPattern daily
One email a day at 7 AM. Sources and a take on every story. Unsubscribe anytime.