GPT-5.4 Pro

GPT-5.4 Pro launched on 5 March 2026 alongside GPT-5.4 and carries the same 1,050,000-token context window, 128,000 tokens of output, text and image input, and a knowledge cutoff of 31 August 2025. Reasoning effort runs medium (the default), high and xhigh — there is no “none” setting, which is the point of the tier. It costs $30 per million input tokens and $180 per million output, twelve times the base model on input and twelve on output. The long-context surcharge applies here too: cross 272,000 input tokens and the entire session is billed at twice the input rate and 1.5 times the output rate. At that point a single million-token prompt costs $60 in input alone. Regional processing endpoints for data residency add a further 10 per cent. OpenAI offers it through the Responses API only, without batch support, and recommends background mode because individual requests can run for several minutes. Its own comparison table places it against o3-pro at $20 and $80 — the previous generation of the same idea, at two thirds the input price.

▸Videos

No videos yet.

▸Reviews