%20--%3e%3cdefs%3e%3cstyle%3e%20.st0%20{%20fill:%20%23061b40;%20}%20.st1%20{%20fill:%20%23306af1;%20}%20.st2%20{%20fill:%20%235ce5cf;%20}%20%3c/style%3e%3c/defs%3e%3cg%3e%3cpath%20class='st0'%20d='M55,10.5h9v3h-9v9h12V7.5h-12v3ZM64,19.5h-6v-3h6v3Z'/%3e%3cpolygon%20class='st0'%20points='69%2016.5%2078%2016.5%2078%2019.5%2069%2019.5%2069%2022.5%2081%2022.5%2081%2013.5%2072%2013.5%2072%2010.5%2081%2010.5%2081%207.5%2069%207.5%2069%2016.5'/%3e%3cpolygon%20class='st0'%20points='95%2010.5%2095%207.5%2083%207.5%2083%2022.5%2095%2022.5%2095%2019.5%2086%2019.5%2086%2016.5%2095%2016.5%2095%2013.5%2086%2013.5%2086%2010.5%2095%2010.5'/%3e%3cpath%20class='st0'%20d='M40,1.5v21h11.6l1.4-1.4v-7.6h0c0,0-1.4-1.5-1.4-1.5l1.4-1.4V2.9l-1.4-1.4h-11.6ZM50,19.5h-7v-6h7v6ZM50,10.5h-7v-6h7v6Z'/%3e%3c/g%3e%3cpath%20class='st1'%20d='M23.1,24L14.7,4.8l-4.9,11.2h3.8l-1.8,4H3.3L12.1,0H2C.9,0,0,.9,0,2v20c0,1.1.9,2,2,2h21.1Z'/%3e%3cpath%20class='st2'%20d='M34,0h-16.8l10.6,24h6.2c1.1,0,2-.9,2-2V2C36,.9,35.1,0,34,0ZM32.5,20h-4V4h4v16Z'/%3e%3c/svg%3e)
Join Now
Microsoft officially launched the lightweight variant of its self-developed image generation model, MAI-Image-2.6-Flash, in September and has opened public preview through Microsoft Foundry. As a high-efficiency version of the MAI-Image-2.6, which topped the technical ranking last month, the release of this model marks Microsoft's comprehensive expansion into industrial-level scenarios that require high concurrency, low latency, and cost sensitivity in the field of AI image generation.
Test data shows that the image generation speed of MAI-Image-2.6-Flash is 2.8 times that of GPT-Image-2-Medium, with an overall comprehensive efficiency improvement of 72%. While significantly reducing inference latency and computational costs, this variant fully inherits the core generation and editing capabilities of the main model, supporting high-precision text-to-image generation and local object-level image editing.
In addition, the entire MAI-Image-2.6 series brings multiple architectural upgrades, including support for up to five reference images to ensure consistency of characters and products across multiple images, integration of Bing search's web positioning function to supplement real-world image information, and new native support for dynamic aspect ratios and higher resolution outputs.
In terms of commercial pricing strategy, the Flash version has reduced the price per million tokens for text input, image input, and image output to 1.75 dollars, 2.50 dollars, and 19 dollars respectively, achieving a significant cost reduction of more than 50% compared to the main model.
Microsoft recommends using the main model for commercial design and final production assets that require stringent image quality and text rendering, while precisely positioning the Flash version for interactive applications, rapid creative iteration, and large-scale automation processes. This multimodal model matrix that balances extreme performance with optimal cost reflects the industry trend of generative AI evolving from single technological breakthroughs to high-throughput, low-cost engineering implementation.