PHMAI-Image-2 is Microsoft’s new in-house text-to-image model, focused on photorealism, reliable text in images, and complex scene generation, and it’s starting to roll out across Copilot, Bing Image Creator, and the MAI Playground.[1][2]
What MAI-Image-2 is
- Second‑generation Microsoft image model, built by the MAI “superintelligence” team, not an OpenAI white‑label.
- Currently ranked around #3 on the You do not have permission to view the full content of this post. Log in or register now. text‑to‑image leaderboard, behind Google and OpenAI image models.
- Designed with input from photographers, designers, and visual storytellers to better match creative workflows.
Key capabilities
- Enhanced photorealism: more natural lighting, better skin tones, and environments with realistic texture and wear, so less post‑processing is needed.
- Strong in‑image text: does signs, posters, infographics, slides, and diagrams with more legible, controllable lettering than prior models.
- Complex scenes: handles dense compositions, cinematic framing, surreal scenes, and detailed multi‑object layouts.
Example use: generating a marketing poster with realistic people plus accurate headline/body text directly from a prompt, instead of compositing text in a separate design tool.
Access and rollout
- Available now in the MAI Playground for users in supported regions like the US.
- Rolling into Copilot and Bing Image Creator, gradually replacing or sitting alongside earlier image backends.
- A developer API and broader commercial use are planned; early commercial use may require an application or approval.
How it compares today
| Aspect | MAI-Image-2 (Microsoft) | Google Gemini image models | OpenAI GPT Image 1.x* |
| Leaderboard spot | Around #3 on Arena.ai | Above MAI-Image-2 on Arena | Above MAI-Image-2 on Arena |
| Strengths | Photorealism, text, scenes | Quality + flexible pricing | Overall fidelity, editing tools |
| Integration | Copilot, Bing, MAI Playground | Gemini apps, Imagen API | ᑕᕼᗩTGᑭT, OpenAI API |
Safety and constraints
- Uses strict safety filters and content restrictions, especially around sensitive and potentially harmful content.
- Positioned as a production‑grade model for enterprise and consumer products, so outputs are tuned more for reliability and policy compliance than “anything goes” experimentation.
Test it for FREE now!
Examples:
MAI-Image-2
Nano Banana Pro
Prompt 1
Code:
A mesmerizing, high-contrast cinematic shot captures the raw energy and profound elegance of a live band performing on stage, bathed in a symphony of dramatic backlighting. The musicians, reduced to striking, sharply defined silhouettes, stand against a mesmerizing aurora of intensely vibrant stage lights – electric blues, fiery reds, and deep purples – diffused through a haze of performance smoke that accentuates the light beams. A low-angle, wide-lens perspective, reminiscent of iconic concert photography, emphasizes their powerful, larger-than-life forms: the lead singer's microphone held aloft, the drummer's arms a blur of motion, the guitarist's iconic stance, each contour etched with a subtle, glowing rim light that separates them from the luminous background. This dynamic composition and moody, high-contrast aesthetic create an unforgettable visual, celebrating the abstract beauty of form and light where music transcends the visual, conveying pure passion and an iconic moment of collective artistry. --ar 3:2 --v 6.0MAI-Image-2
Nano Banana Pro
Prompt 2
Code:
Create a highly photorealistic ultra-wide panoramic image of a quiet rural prairie landscape at night, featuring a weathered red wooden barn with a silo on the left side of the frame and a tall metal windmill with a water tank on the right, both silhouetted against a vivid Milky Way galaxy stretching diagonally across the sky.
Captured with a professional full-frame mirrorless camera using a 14mm ultra-wide angle lens at f/2.0, ISO 3200, with a 20-second long exposure under clear, moonless night conditions.
The scene takes place in an open grassland prairie with rolling hills in the distance, wooden fence lines leading through the foreground toward both structures, and a small distant farmhouse with warm glowing window lights near the horizon.
The barn is positioned in the left foreground with visible aged wood texture and subtle edge lighting, while the windmill stands tall in the right midground, slightly silhouetted with faint rim lighting from ambient sky glow. The composition uses leading lines from the fence and dirt path to guide the viewer’s eye across the frame.
The Milky Way core is highly detailed, with dense star clusters, interstellar dust lanes, and subtle color variations in warm oranges, cool blues, and faint magentas, naturally blending into the night sky. A faint green airglow is visible near the horizon.
Foreground grasses show slight motion blur from a gentle breeze, while all celestial elements remain sharp due to careful exposure timing.
The image must contain authentic real-world photographic imperfections such as:
- subtle high-ISO sensor noise in darker areas
- slight vignetting toward the corners
- natural star diffraction and minor chromatic aberration
- realistic long-exposure light falloff
- faint atmospheric haze near the horizon
Environmental lighting should include:
- soft ambient illumination from starlight
- gentle skyglow providing minimal visibility to terrain
- realistic shadow gradients with no artificial light sources except the distant house
Materials and textures must be physically accurate, including:
- rough wood grain on the barn
- metallic reflections on the windmill structure
- natural grass variation and density
- dusty rural terrain with small rocks along the path
Colors must remain true-to-life with a balanced astrophotography color grade, avoiding oversaturation or artificial contrast.
Ensure correct perspective distortion from the ultra-wide lens, with natural horizon curvature and spatial depth, making the scene feel expansive and immersive.
The final image should be indistinguishable from a professionally captured astrophotography landscape photo, with cinematic yet realistic composition and lighting.MAI-Image-2
Nano Banana Pro
Prompt 3
Code:
Create a highly photorealistic image of a field of blooming pink cosmos flowers captured with a professional full-frame mirrorless camera using a wide-angle 24mm lens at f/2.8 in bright natural daylight.
The scene takes place in an open meadow under a vivid blue sky with soft, wispy cirrus clouds stretching across the upper frame.
The flowers are tall and slender, with delicate pink petals and thin green stems, gently leaning and swaying as if in a light breeze. The camera is positioned very low to the ground, looking upward, creating a dramatic perspective where the flowers tower toward the sky.
The composition places the denser cluster of flowers on the right side of the frame, while the left side opens into expansive sky, giving a sense of depth and airiness. Foreground flowers are in sharp focus while background flowers gradually blur into a soft bokeh.
The image must contain authentic real-world photographic imperfections such as:
- subtle lens distortion from the wide-angle perspective
- natural sensor noise in shadow areas
- optical depth-of-field falloff
- realistic sunlight highlights and soft shadows on petals
- fine details like pollen, petal veins, and tiny imperfections
- slight chromatic aberration along high-contrast edges
Environmental details should include:
- gentle atmospheric haze near the horizon
- natural sky gradient from deeper blue at the top to lighter near the horizon
- sunlight interacting with petals creating slight translucency
- subtle motion blur on a few distant flowers
Colors must remain physically accurate with balanced white tones and natural, slightly warm daylight grading.
Ensure accurate:
- botanical structure of cosmos flowers
- natural randomness in flower spacing and orientation
- realistic stem curvature and overlapping layers
- soft wind-influenced positioning
Camera behavior must simulate real optics including:
- perspective exaggeration from low-angle wide lens
- shallow depth of field in foreground-to-background transition
- natural framing with slight asymmetry
Include natural imperfections like:
- minor petal tears or bends
- uneven bloom sizes
- small insects or specks on some flowers
- slight dust particles catching sunlight
The final image should be indistinguishable from a photograph captured by a skilled outdoor nature photographer and must obey real-world physics and lighting behavior.UPDATE
Microsoft just dropped MAI-Image-2-Efficient — and it's a game changer for creators and businesses.
22% faster image generation 41% cheaper than the flagship model 4x more compute-efficient Available NOW — no waitlistWhether you're generating product shots, marketing assets, or UI mockups at scale, this is your new go-to model. Microsoft is calling it the "production workhorse" — and the numbers back it up.
Find it on Microsoft Foundry and MAI Playground today.
Learn more about it here: You do not have permission to view the full content of this post. Log in or register now.
Your feedback is highly appreciated
Support my other posts
- Google just KILLED Photoshop!
- 50 Brilliant Ways to Supercharge Creativity with Nano Banana
- Nano Banana Prompt Gallery
- AI Fashion Studio: AI Virtual Try-On Powered By Nano Banana
- Free Image Upscaler up to 16K Quality!
- Travel the World with Nano Banana
- Nano Banana Polaroid Trend
- AI Profile Picture Generator
- AI Snapshot Generator
- ᑕᕼᗩTGᑭT Prompt Packs
- Perplexity at Work
- Free AI Image Editor
- DumPDF: PDF Editor
- LuxPDF: Open Source PDF Tools
- Affinity Studio: Free, Powerful Design Tool
- Gemini Edu ID Card Generator
- CanVâ Education Invite Link 2
- Create UNCENS0RED/NSFW AI Characters
- Student ID Card Prompt
- Introducing Nano Banana Pro
- Nano Banana Pro Image And Prompt Gallery
- Create 4K Nano Banana Pro Images
- Create Pro-Grade Infographics
- IHatePDF: Toolkit For Everyday Documents
- OpenClaw: An AI Agent That Actually Does Things
- Stunning Nano Banana Prompts Gallery
- Lyria 3: Google's AI Music Studio
- Meet Gemini 3.1 Pro
- Create City Map Posters
- Seedream 5.0 Lite: A Smart, Web-Aware AI Image Model
- Nano Banana 2: ProLevel Image Generation at Flash Speed
- Meet Gemini 3.1 Flash‑Lite: Google’s New High‑Throughput AI Workhorse
- GPT‑5.3 Instant: Smarter, Faster Everyday Chat
- GPT‑5.4: OpenAI’s New Flagship GPT‑5‑Series Model