OpenAI restores the five-hour Codex and ChatGPT Work limit for Plus accounts
On August 25, 2026, the Codex and ChatGPT Work usage page again showed a five-hour window for Plus users; users were still reporting the limit in effect on August 28.
EVENT ARCHIVE
Events
20 English reports are shown on this page. Refine the keywords, date range, company, or event type below.
AI-generated · awaiting review
NVIDIA says SpaceXAI will adopt Vera CPUs for Grok infrastructure
NVIDIA announced that SpaceXAI will deploy Vera CPUs and expand Grok infrastructure with Vera Rubin; the companies also plan an optimized Vera Rubin NVL72 system for the first-generation Starmind AI satellite.
Codex full paid-plan reset propagated to accounts
Tibo said the full reset previously promised for all paid Codex subscriptions had been propagated to accounts, and that some fixes for the usage issues found the previous day had landed.
Codex 20M milestone banked reset delivered
After Codex reached 20M active users, Tibo confirmed that a banked reset had landed for all paid ChatGPT Work and Codex users.
OpenAI cuts GPT-5.6 Sol API pricing
OpenAI lowered GPT-5.6 Sol Standard short-context pricing to $4 per million input tokens and $20 per million output tokens, down 20% and 33%; the promotional pricing lasts at least through November 21, 2026.
Mistral launches Agentic Search
Mistral launched Agentic Search, a multi-step retrieval loop that lets models search, open, navigate, read, and verify complex documents through Search Toolkit and Libraries.
Qwen releases the open-source Qwen3.8-27B
Qwen releases Qwen3.8-27B, a compact open-source model with visual understanding, adjustable thinking depth and long-horizon agent capabilities.
Z.ai releases GLM-5.3
Z.ai releases GLM-5.3 with stronger long-horizon agent, coding and automated vulnerability-discovery capabilities.
Anthropic announces text watermarking for future Claude models
Anthropic announced that future Claude models will add an invisible text watermark to meet EU AI Act content-marking requirements, without adding tokens, cost, or user-identifying information.
OpenAI previews Cerebras-powered GPT-5.6 Sol Ultrafast
OpenAI opens a limited GPT-5.6 Sol Ultrafast preview to selected API customers, powered by Cerebras and generating up to 750 output tokens per second.
Qwen releases the open Qwen3.8-2.4T-A95B
QwenCloud lists Qwen3.8-2.4T-A95B as the open version of the flagship Qwen3.8 family.
Google releases Gemini 3.7 Flash
Google DeepMind releases Gemini 3.7 Flash, an algorithmic advance over Gemini 3.6 Flash with text, image, audio and video input, text output and up to one million tokens of context. Gemini API documentation lists it as a stable Gemini 3 model.
Codex 15M milestone usage reset
After Codex crossed 15 million active users, Tibo announced a usage reset and said it would land within about an hour.
DeepSeek updates the deepseek-v4-pro API to DeepSeek-V4-Pro-0813
DeepSeek official documentation confirms that deepseek-v4-pro now serves DeepSeek-V4-Pro-0813 under the unchanged API model name, with a one-million-token context window, up to 384K output tokens, and Responses API support.
xAI releases Grok 4.6
xAI releases Grok 4.6 for long-running agents, coding and knowledge work, with a 500K-token context window.
Promised Monday reset completed
Tibo confirmed that the promised Monday Codex reset was done; the completion post did not restate its scope or reset form.
OpenAI tests paid Codex usage resets with a subset of users
OpenAI shows some ChatGPT Plus and Pro users a paid reset option after their Codex allowance is exhausted. Public reports cite about $8 for Plus and about $40 or $80 for Pro 5x and 20x, but eligibility, prices and presentation vary by account and OpenAI has not published a unified reset price list.
Meta releases Muse Glimmer 30B
Meta publicly releases the Muse Glimmer 30B model weights and related inference components.
OpenAI resets paid-user limits to celebrate GPT-5.6 Sol in Claude Code
Tibo said GPT-5.6 Sol can be used nearly anywhere, including the Claude Code (CC) harness, and, while also noting that he would not leave, confirmed a usage-limit reset for all paid ChatGPT Work and Codex users.
LG AI Research releases the open-weight K-EXAONE 2.0
LG AI Research publishes the K-EXAONE 2.0-750B-A37B model weights and technical report.