CodexHowSupport Us

GPT-5.4 Cyber — Specs, Pricing & Limits

Specs

StatusNot published
Context windowNot published
Max outputNot published
Knowledge cutoffNot published
Reasoning effort levelsNot published
Long-context ruleNot published

Pricing (short context)

PricingNot published

Rate limits by usage tier

Rate limitsNot published

GPT-5.4 Cyber is listed by name in the raw model-id enumeration on OpenAI's own pricing page source, which is the only reason this site's research ever found it — it doesn't appear in the rendered pricing table with an actual price, and its own dedicated reference page returns nothing when fetched directly.

Every field this site would otherwise publish for a model — context window, maximum output, knowledge cutoff, rate limits, every price cell across every service tier — is unavailable for this entry specifically, not defaulted to zero or inherited from a sibling model. That distinction matters: this isn't a free or trial-tier model, it's one where the underlying information simply hasn't been made public yet, on either surface this site checked.

Given the "Cyber" naming and its position in the GPT-5.4 generation's id lineup, it's a reasonable guess that this model targets some specialized use case rather than general availability — but this site doesn't publish guesses as facts, and nothing about that naming pattern is confirmed by either checked source. It is not selectable in any calculator on this site, and won't become selectable until OpenAI actually publishes a price for it.

Verified 2026-08-09 against https://developers.openai.com/api/docs/pricing.

Could not confirm: Every price cell and every model-page field is unpublished. The model page itself 404s.

Checked: https://developers.openai.com/api/docs/pricing · https://developers.openai.com/api/docs/models/gpt-5.4-cyber