GPT Image 1.5

GPT Image 1.5 is the current flagship image model from OpenAI, launched in March 2026 as the successor to GPT Image 1. The model is multimodal — accepting both text and image inputs to produce high-fidelity outputs — and represents a significant leap in generation speed (4x faster), prompt adherence, and text rendering accuracy. The model introduces token-based billing for internal reasoning, making complex multi-element generations more transparent in cost. Image tokens are approximately 20% cheaper than the previous GPT Image 1, offsetting the new reasoning token costs for most use cases. Precise editing via masks allows users to modify specific regions of an image without affecting the rest. GPT Image 1.5 is available directly in ChatGPT for subscribers and via the OpenAI API with tiered pricing by quality: low ($0.009), medium ($0.034), or high ($0.133) at 1024x1024 resolution. Portrait and landscape orientations are priced slightly higher.