OpenAI resets Codex and ChatGPT Work usage limits

// 1h agoPRODUCT UPDATE

OpenAI resets Codex and ChatGPT Work usage limits

OpenAI reset usage limits for Codex and ChatGPT Work, extending GPT-5.6 Sol capacity by roughly 18%. The temporarily suspended five-hour limit returns tomorrow, though overall quality remains unaffected.

// ANALYSIS

A welcome capacity boost for power users, extending the utility of the latest models while reintroducing necessary throttling mechanisms.

–Usage limits have been reset across Codex and ChatGPT Work
–GPT-5.6 Sol usage is extended by roughly 18%
–The suspended five-hour limit returns tomorrow
–OpenAI confirmed actual limits were never reduced and quality remains unaffected

// TAGS

openaicodexchatgptgpt-5.6-solllmai

DISCOVERED

1h ago

2026-07-29

PUBLISHED

2h ago

2026-07-29

RELEVANCE

8/ 10

AUTHOR

mark_k

// KEEP READING

More AI developer news from the feed

EXPLORE FULL FEED

UPDATE28m ago

xAI releases Grok Voice Think Fast 2.0

SpaceXAI has announced a major upgrade to Grok Voice with the release of Grok Voice Think Fast 2.0. The updated speech-to-speech model tops the Artificial Analysis Speech-to-Speech Quality Index with an 82.9% rating, reduces time to first audio from 1.25s down to 0.70s, and delivers major transcription improvements in noisy real-world environments.

MODEL34m ago

Claude 5 Opus requires new interaction paradigms

A post by AI researcher Omar Sanseviero discusses early impressions of Claude 5 Opus, noting that the model and the broader Claude 5 family are trained to operate with significantly higher agency than previous models. This shift in model capabilities requires users to rethink standard prompting strategies and contextualization methods when interacting with Opus 5.

OPEN SOURCE36m ago

Shared-Claude-Chats archives public Claude and Grok chats

Shared-Claude-Chats is an open-source repository that archives publicly shared Claude and Grok conversation links into markdown files alongside scraping scripts. The project creates a dataset of real-world AI prompt-response pairs while highlighting the privacy risks of shared chat links.