Mira Murati

New Capabilities

Public face of GPT-5's technical capabilities

OpenAI's GPT-5 generation models, GPT-5.4 and GPT-5.5, accept up to one million tokens per request. That is enough for a full codebase or an entire book, and about 125 times GPT-4's 8,000-token window. The base GPT-5 API model takes 272,000 input tokens with a 400,000-token total context.

Updated Oct 26

New Capabilities

As of May 2026: Leading gigawatt-scale Vera Rubin deployment partnership with Nvidia

Nvidia has spent four years on an annual architecture cadence that no semiconductor company has sustained before. At GTC 2026, Jensen Huang unveiled Vera Rubin (a single-GPU system delivering 50 petaflops, five times Blackwell's performance, at one-tenth the cost per token) and NemoClaw, an open-source platform letting companies deploy autonomous AI agents without cloud-provider lock-in.

Updated May 30