GPT-5.5 Pro Spotted in Codex App Settings

GPT-5.5 Pro, OpenAI's top-tier reasoning model currently restricted to ChatGPT, appears headed to the Codex coding agent. On June 11, X user @argofowl discovered a 'GPT 5.5 Pro' entry within the Codex app settings and shared a screenshot of the sighting.

While the model has not yet been added to the active selection menu and OpenAI has made no official announcement, the appearance of a billing entry suggests a public rollout is imminent.

Unannounced Model Appears in Billing Settings

The sighting came from the 'Usage & Billing' search in the Codex app settings. The original post shows the GPT 5.5 Pro entry positioned alongside currently supported models.

AI engineer Mikel (@MikelEcheve) confirmed the visibility of the entry. 'Not announced. Not available yet. But if it's already showing up in the app, OpenAI may be getting ready to roll it out soon,' he noted.

This is not the first time upcoming models have left traces in the coding agent before their official debut. In late May, a routing string for 'gpt-5.6' briefly appeared in internal logs before being removed, and an unreleased model name surfaced on the pricing page last month. The recurring pattern of settings-level leaks followed by official launches suggests a similar rollout timeline for the Pro model.

A Highly Anticipated Upgrade for Developers

GPT-5.5 Pro debuted alongside the standard GPT-5.5 model on April 23 but has since been restricted to ChatGPT Pro, Business, and Enterprise tiers. Users of the coding agent currently have access only to the standard GPT-5.5 model, which features a 400,000-token context window.

The Pro variant allocates additional compute at inference time to achieve higher accuracy. Its API pricing is set at $30 per million input tokens and $180 per million output tokens, representing a sixfold increase over the standard version.

Consequently, developers have frequently petitioned for Pro model support within the Codex command-line interface. Many have relied on workarounds, exporting complex software architecture and debugging tasks to ChatGPT Pro—a friction point that explains why the sighting set off such a strong community reaction.

Escalating Competition in Coding Agents

The strategic rationale for the integration is straightforward. Codex, which recently surpassed five million weekly active users, is locked in intense competition with Anthropic's Claude Code. Embedding its premier reasoning model directly into the coding agent is a logical progression in this rivalry.

The primary uncertainty remains the scope of the rollout. It is still unclear which tiers the rollout will cover and whether OpenAI will impose specific usage limits to manage the model's high computational costs.

However, if historical patterns persist, a formal announcement could follow the sighting within days.