Microsoft AI today rolled out MAI-Image-2.6-Flash, a faster version of its flagship MAI-Image-2.6 model, and it is now in public preview on Microsoft Foundry. It keeps the same core features as the original: text-to-image generation, image editing, multi-image reference, web search for latest information, and automatic aspect ratio selection. The big idea is simple—cut latency and lower cost.
Speed Boost: 2.8x Faster, 72% More Efficient
Speed is the main draw here. MAI-Image-2.6-Flash produces images 2.8 times faster than GPT-Image-2 Medium and uses GPU resources 72% more efficiently. But the extra speed is not coming from a drop in quality. According to independent benchmark platform Artificial Analysis, the model posts a 1311 Elo score in image editing, which puts it in third place and just ahead of GPT Image 2 high at 1309 Elo. For text-to-image generation, it sits eighth with 1297 Elo. And the flagship MAI-Image-2.6 still leads overall on raw strength, holding the No. 1 position in image editing and No. 2 in text-to-image right now.
Pricing: Under $20 per 1,000 Images
As for price, Microsoft Foundry lists MAI-Image-2.6-Flash outputs at $19 per million tokens, versus $38 per million tokens for the flagship MAI-Image-2.6. Artificial Analysis puts the cost at roughly $19.5 per 1,000 images for Flash. GPT Image 2 high, by comparison, comes in at about $211 per 1,000 images. Big gap. That makes Flash a much more cost-effective choice.