gpt-5.6-sol has already appeared in the models.json repository.

The context size I care most about is also basically confirmed: 372K. Compared to the previous generation gpt-5.5's 272K, that's an improvement of about 36.7%.

But to be honest, seeing this number is still a bit disappointing. After all, opus 4.8 and fable-5 have already achieved 1M context, so 372K doesn't really seem that impressive by today's standards.

Still, Codex's compression ability has always been strong, and stability remains good after multiple rounds of compression in long tasks. So in actual use, the impact may not be as significant as the numbers suggest.

Overall, it's a steady upgrade, but I still hope the next generation will max out the context window.