OpenAI's flagship reasoning model — third on GPQA-Diamond at 93.5 — and the model behind the recent enterprise pricing reset (2× per-token vs GPT-5.4) plus the Codex CLI subscription bundle.
The 2× pricing move signals OpenAI's confidence that frontier reasoning has pricing power even as open-weight competitors close the score gap. For enterprise budgets, the spread between Sonnet 4.6 and GPT-5.5 just widened.
The Codex absorption simplifies billing but forces a migration for teams that built on the standalone Codex endpoint — there's a 20% effective price hike for migrators that the Codex CLI subscription partially offsets.
CTOs renegotiating frontier-model spend, product engineers using Codex for codegen, and infra teams planning AWS-vs-Azure-vs-GCP placement.