Microsoft has officially rolled out its new in-house image generation model, MAI-Image-1, for all users on Bing.com/create and the Bing mobile app. The new model is part of Microsoft’s growing lineup of in-house AI systems designed to provide faster, more efficient, and more controlled AI experiences across its platforms.
While Microsoft continues to have full access to OpenAI’s latest models, the company is also investing heavily in developing its own AI models. Earlier this year, it introduced MAI-Voice-1 and MAI-1-preview, and now it has followed up with MAI-Image-1, which is a direct competitor to OpenAI’s gpt-image-1 and Google’s Gemini/Imagen models.
This move underlines Microsoft’s strategy to create a differentiated AI ecosystem, blending OpenAI technologies with its own innovations.
Users who want to try the new model can find it on Bing’s image generator page. They will need to select MAI-Image-1 from the model picker before generating images. The feature is available in all countries where Bing Image Creator and Copilot Labs operate, with Microsoft promising that EU access will be added soon.
According to Microsoft’s AI team, the MAI-Image-1 model was trained with special attention to reducing repetitive or overly stylized outputs. The team used a rigorous data selection and task-based evaluation process to ensure the model produces more realistic and diverse images.
The company also highlighted that MAI-Image-1 delivers faster performance compared to many other large image-generation models. This improvement could make Bing’s image generation experience smoother and more responsive, even during heavy use.
