GPT-6.1 Sol's default context in Codex and Work goes from 272k to 500k, with no long-context fee on subscriptions
What is it?
On day 7 of its "28 days of improvements" run, OpenAI raised GPT-6.1 Sol's default context (how much input the model holds at once) in Codex and ChatGPT Work from 272k to 500k tokens, roughly an 80% increase. Users can still set it anywhere up to 1M tokens. When usage comes from a subscription, long-context fees above 272k are waived too.
Why does it matter?
OpenAI says long tasks should see 2-3x fewer compactionsVendor claim, the step where the conversation gets summarized to fit. Agents working across big codebases should get interrupted less and lose less nuance. It follows the Ultrafast mode we covered earlier. This time the change is in how much the model remembers rather than how fast it answers.
Maturity
GAThis only applies to the Codex and Work subscription path. The API model docs still list long-context pricing (2x input) above 272k, and community members say the Codex CLI may need a release with updated model config before the new default takes effect.
How to track it
- Whether the API gets 500k and new long-context pricing: check the above-272k pricing note in the GPT-6.1 Sol model docs.
- Codex CLI support: watch
openai/codexGitHub releases for a version with the updated model config. - The next "28 days" improvements: follow the OpenAI community thread and Tibo (@thsottiaux) on X.