Microsoft AI 发布 MAI Image 2.7 图像编辑模型，在 Arena 和 Artificial Analysis 取得前列排名

Our Company

Signal blog Official Microsoft Blog Command Line Microsoft On The Issues Asia Canada Europe, Middle East and Africa Latin America The Code of Us

What's new AI Innovation Digital Transformation Sustainability Security Work & Life Diversity & Inclusion Unlocked

Our CompanySignal blog Official Microsoft Blog Command Line Microsoft On The Issues Asia Canada Europe, Middle East and Africa Latin America The Code of Us

What's new AI Innovation Digital Transformation Sustainability Security Work & Life Diversity & Inclusion Unlocked

All Microsoft

Software

PCs & Devices

Entertainment

Business

Developer & IT

Other

View Sitemap

Cancel

Models ## Latest models

Discover the latest MAI models, designed for real-world intelligence

Image 2: A watercolor-style shield with soft pastel colors blending yellow, pink, and green on a plain white background. ### MAI-Cyber-1.1-Flash Image 3: Two large watercolor curly brackets face each other on a white background. The left bracket is blue and green, while the right bracket is pink and orange. ### MAI-Code-1.1-Flash Image 4: Two watercolor speech bubbles overlapping on a light background; one bubble is pink and yellow, the other is green. ### MAI-Voice-2.1 Image 5: A watercolor illustration of a soundwave with soft, overlapping peaks in green, pink, and yellow tones on a light gray background. ### MAI-Transcribe-2 Image 6: A watercolor illustration of a thought bubble made up of overlapping pastel circles in pink, yellow, green, and blue, set against a light gray background. ### MAI-Thinking-1 All models Image 7: A watercolor-style shield with soft pastel colors blending yellow, pink, and green on a plain white background.

MAI-Cyber-1.1-Flash

A new cybersecurity model built into MDASH, our multi-agent vulnerability identification and remediation harness. Image 8: Two large watercolor curly brackets face each other on a white background. The left bracket is blue and green, while the right bracket is pink and orange.

MAI-Code-1.1-Flash

Helps your engineering team write better code faster, with a lightweight, agentic model built into GitHub Copilot and VS Code. Image 9: Two watercolor speech bubbles overlapping on a light background; one bubble is pink and yellow, the other is green.

MAI-Voice-2.1

Delivers expressive, low-latency speech that holds up across long generations. Image 10: A watercolor illustration of a soundwave with soft, overlapping peaks in green, pink, and yellow tones on a light gray background.

MAI-Transcribe-2

Turn noisy audio into precise, domain-specific transcripts, with leading FLEURS and Artificial Analysis accuracy scores. Image 11: A watercolor illustration of a thought bubble made up of overlapping pastel circles in pink, yellow, green, and blue, set against a light gray background.

MAI-Thinking-1

Built for the complex problems that matter most, with competitive reasoning and top SWE-Bench Pro results at a mid-weight price.

Blog ## Featured stories

The latest breakthroughs and updates from within our team

Image 12: Blurry, light-colored flowers against a blue sky. ## MAI Image 2.7 debuts above GPT Image 2.5 Flare on Artificial Analysis and #3 on Arena

2 min read Image 13: A shield-shaped icon with a watercolor texture in pastel yellow, pink, and blue against a soft, pink, flowing background. ## MAI Cyber 1.1 Flash: Sharper findings and fewer tokens inside MDASH

2 min read Image 14: A white heart icon composed of rounded square dots, centered against a blurred, colorful floral background. ## Building Medical Superintelligence

6 min read All stories

THE HUMANIST REVIEW OF AI

Off- [x] Accessibility Mode

Models

MAI Image 2.7 debuts above GPT Image 2.5 Flare on Artificial Analysis and #3 on Arena

MAI Image 2.7 debuts above

GPT Image 2.5 Flare on

Artificial Analysis and #3 on

Arena

October 27, 2026

Models

Image 15: Blurry, light-colored flowers against a blue sky.

MAI Image 2.7, our latest image editing model, launches with even stronger control for edits that are hardest to get right. It launches in preview at No. 3 on Arena Edit, making Microsoft AI the No. 2 lab, and No. 2 on Artificial Analysis’s image editing leaderboard.

Restyle or relight an entire scene, change text and symbols while keeping them integrated into the image, or restore and enhance existing assets. And make edits that require understanding how objects, lighting, and environments should work together.

<meta http-equiv="Content-Security-Policy" content="default-src 'none'; script-src 'unsafe-inline' 'unsafe-eval'; style-src 'unsafe-inline'; img-src data: blob: https://microsoft.ai/wp-content/uploads/; font-src data: https://microsoft.ai; media-src data: blob: https://microsoft.ai/wp-content/uploads/;"><div class="mai-elo" id="maiElo">
  <header class="el-hd">
    <h2 class="el-title">MAI-Image-2.7-Preview ranks 2nd of all public models</h2>
    <p class="el-sub">Elo from blind preference votes in the Image Arena</p>
  </header>

  <div class="el-plot" id="elPlot" role="img" aria-label="Horizontal bar chart of Image Editing Arena Elo. GPT Image 2.5 Sunburst (max) 1199, MAI-Image-2.7-Preview 1193, GPT Image 2.5 Flare (max) 1182, MAI-Image-2.6 1151, Krea 3 (Ultra) 1151, Nano Banana 2.1 1151, MAI-Image-2.6-Flash 1144, GPT Image 2 (high) 1136."></div>

  <ul class="el-key" aria-hidden="true">
    <li><i class="el-sw el-sw-mai"></i><em>MAI-Image-2.7-Preview</em></li>
    <li><i class="el-sw el-sw-oth"></i>MAI-Image-2.6 family</li>
  </ul>

  <p class="el-src">Source: Artificial Analysis &middot; AA-Image-Editing v2.1</p>
</div>

<style>
@font-face{font-family:"Bradford LL";font-weight:450;font-style:normal;font-display:swap;src:url("https://microsoft.ai/wp-content/themes/wp-base-theme/dist/assets/BradfordLLWeb-Regular-C6SjiFc8.woff2") format("woff2");}@font-face{font-family:"Bradford LL";font-weight:450;font-style:italic;font-display:swap;src:url("https://microsoft.ai/wp-content/themes/wp-base-theme/dist/assets/BradfordLLWeb-Italic-CA7el69X.woff2") format("woff2");}@font-face{font-family:"Bradford LL";font-weight:600;font-style:normal;font-display:swap;src:url("https://microsoft.ai/wp-content/themes/wp-base-theme/dist/assets/BradfordLLWeb-Bold-BJqvqpMA.woff2") format("woff2");}@font-face{font-family:"Bradford LL";font-weight:600;font-style:italic;font-display:swap;src:url("https://microsoft.ai/wp-content/themes/wp-base-theme/dist/assets/BradfordLLWeb-BoldItalic-B6a2GBYT.woff2") format("woff2");}@font-face{font-family:"Red Hat Mono";font-weight:400;font-style:normal;font-display:swap;src:url("https://microsoft.ai/wp-content/themes/wp-base-theme/dist/assets/RedHatMono-Regular-CEF4ave-.woff2") format("woff2");}@font-face{font-family:"Red Hat Mono";font-weight:500;font-style:normal;font-display:swap;src:url("https://microsoft.ai/wp-content/themes/wp-base-theme/dist/assets/RedHatMono-Medium-BxJECwMl.woff2") format("woff2");}
html,body{margin:0;}
body{background:transparent;}
.mai-elo,.mai-elo *,.mai-elo *::before,.mai-elo *::after{box-sizing:border-box;}

.mai-elo{
  font-family:"Bradford LL", Georgia, serif;
  font-weight:400;
  color:#5D524B;
  -webkit-font-smoothing:antialiased;
}

.el-hd{margin:0 0 40px;}
.el-title{
  font-family:"Bradford LL", Georgia, serif;
  font-weight:400;font-size:32px;line-height:1.2;letter-spacing:-0.02em;margin:0;color:#5D524B;
}
.el-sub{
  font-family:"Red Hat Mono",ui-monospace,Menlo,Consolas,monospace;
  font-weight:400;font-size:14px;line-height:1.5;margin:10px 0 0;color:#8A8077;
}

/* ---- plot ---- */
.el-row{
  display:grid;
  grid-template-columns:230px minmax(0,1fr) 62px;
  align-items:center;
  column-gap:16px;
  margin-bottom:14px;
}
.el-name{
  font-family:"Bradford LL", Georgia, serif;
  font-weight:400;font-size:17px;line-height:1.25;text-align:right;color:#5D524B;
}
.el-row.is-mai .el-name{color:#546B5C;font-style:italic;}
.el-track{position:relative;height:30px;}
.el-bar{
  position:absolute;left:0;top:0;height:100%;width:0;
  background:#99BFBF;border-radius:0 5px 5px 0;
  opacity:.5;
  transition:width 1s cubic-bezier(0.43,0.195,0.02,1),
             opacity .35s cubic-bezier(0.43,0.195,0.02,1);
}
/* MAI reads at full strength by default; hovering the plot lifts every bar */
.el-row.is-mai .el-bar{opacity:1;}
.el-grid:hover .el-bar{opacity:1;}
.el-row.is-mai .el-bar{background:#738C7B;}
.el-val{
  font-family:"Red Hat Mono",ui-monospace,Menlo,Consolas,monospace;
  font-weight:500;font-size:15px;line-height:1;color:#5D524B;
  font-variant-numeric:tabular-nums;opacity:0;
  transition:opacity .4s cubic-bezier(0.43,0.195,0.02,1);
}
.el-row.is-mai .el-val{color:#546B5C;font-weight:500;}
.is-in .el-val{opacity:1;}

/* gridlines sit behind the bars, axis beneath */
.el-grid{position:relative;}
.el-gl{position:absolute;top:0;bottom:0;width:1px;background:#E3DCCE;}
.el-axis{
  display:grid;grid-template-columns:230px minmax(0,1fr) 62px;
  column-gap:16px;margin-top:4px;
}
.el-ticks{position:relative;height:18px;}
.el-tk{
  position:absolute;top:0;transform:translateX(-50%);white-space:nowrap;
  font-family:"Red Hat Mono",ui-monospace,Menlo,Consolas,monospace;
  font-weight:400;font-size:12px;line-height:1.4;color:#8A8077;
  font-variant-numeric:tabular-nums;
}

.el-key{
  display:flex;flex-wrap:wrap;gap:26px;list-style:none;margin:30px 0 0;padding:0;
}
.el-key li{
  display:inline-flex;align-items:center;gap:9px;
  font-family:"Bradford LL", Georgia, serif;
  font-weight:400;font-size:15px;line-height:1.3;color:#5D524B;
}
.el-sw{display:block;width:14px;height:14px;flex:none;border-radius:3px;}
.el-key em{font-style:italic;color:#546B5C;}
.el-src{
  font-family:"Red Hat Mono",ui-monospace,Menlo,Consolas,monospace;
  font-weight:400;font-size:12px;line-height:1.5;margin:22px 0 0;color:#8A8077;
}
.el-sw-mai{background:#738C7B;}
.el-sw-oth{background:#AFC3B5;opacity:.5;}

/* ---- mobile ---- */
@media (max-width:600px){
  .el-hd{margin-bottom:28px;}
  .el-title{font-size:23px;}
  .el-sub{font-size:12px;margin-top:8px;}
  .el-gl{display:none;}
  .el-row{grid-template-columns:minmax(0,1fr) 54px;column-gap:10px;row-gap:5px;margin-bottom:16px;}
  .el-name{grid-column:1 / -1;text-align:left;font-size:15px;}
  .el-track{height:22px;}
  .el-val{font-size:13px;}
  .el-axis{grid-template-columns:minmax(0,1fr) 54px;column-gap:10px;}
  .el-tk{font-size:10px;}
  .el-key{gap:16px;margin-top:22px;}
  .el-key li{font-size:13px;}
  .el-src{font-size:11px;margin-top:18px;}
}

@media (prefers-reduced-motion:reduce){
  .el-bar,.el-val{transition:none !important;}
}
</style>

<script>
(function(){
  var MIN=1100, MAX=1205;
  var D=[
    ["GPT Image 2.5 Sunburst (max)",1199,0,"#E6CAD1"],
    ["MAI-Image-2.7-Preview",1193,1,"#738C7B"],
    ["GPT Image 2.5 Flare (max)",1182,0,"#FBCFB9"],
    ["MAI-Image-2.6",1151,0,"#AFC3B5"],
    ["Krea 3 (Ultra)",1151,0,"#99BFBF"],
    ["Nano Banana 2.1",1151,0,"#E6CAD1"],
    ["MAI-Image-2.6-Flash",1144,0,"#AFC3B5"],
    ["GPT Image 2 (high)",1136,0,"#FBCFB9"]
  ];
  var TICKS=[1100,1120,1140,1160,1180,1200];
  var MTICKS=[1100,1150,1200];

  var root=document.getElementById("maiElo"), plot=document.getElementById("elPlot");
  if(!root||!plot) return;
  function el(t,c){ var n=document.createElement(t); if(c) n.className=c; return n; }
  function mob(){ return window.matchMedia("(max-width:600px)").matches; }
  var pct=function(v){ return (v-MIN)/(MAX-MIN)*100; };
  var bars=[], vals=[];

  function build(){
    plot.innerHTML=""; bars=[]; vals=[];
    var m=mob();
    var grid=el("div","el-grid");

    D.forEach(function(d,i){
      var row=el("div","el-row"+(d[2]?" is-mai":""));
      var nm=el("div","el-name"); nm.textContent=d[0];
      var tr=el("div","el-track");
      var bar=el("div","el-bar"); bar.style.background=d[3]; bar.style.transitionDelay=(i*70)+"ms";
      tr.appendChild(bar);
      var vl=el("div","el-val"); vl.textContent=MIN; vl.style.transitionDelay=(i*70+260)+"ms";
      row.appendChild(nm); row.appendChild(tr); row.appendChild(vl);
      grid.appendChild(row);
      bars.push([bar,d[1]]); vals.push([vl,d[1]]);
    });
    plot.appendChild(grid);

    var ax=el("div","el-axis");
    if(!m) ax.appendChild(el("div"));
    var tk=el("div","el-ticks");
    (m?MTICKS:TICKS).forEach(function(v){
      var t=el("div","el-tk"); t.textContent=v.toLocaleString();
      t.style.left=pct(v)+"%"; tk.appendChild(t);
    });
    ax.appendChild(tk); ax.appendChild(el("div"));
    plot.appendChild(ax);

    // gridlines behind the bars, aligned to the ticks
    (m?MTICKS:TICKS).forEach(function(v){
      var g=el("div","el-gl"); g.style.left="calc("+pct(v)+"% )";
      grid.appendChild(g);
    });
    positionGrid(grid, m);
  }

  function positionGrid(grid,m){
    // gridlines span the plot column only
    var first=grid.querySelector(".el-track");
    if(!first) return;
    var gr=grid.getBoundingClientRect(), tr=first.getBoundingClientRect();
    var left=tr.left-gr.left, w=tr.width;
    [].forEach.call(grid.querySelectorAll(".el-gl"),function(g,i){
      var v=(m?MTICKS:TICKS)[i];
      g.style.left=(left + w*(v-MIN)/(MAX-MIN))+"px";
    });
  }

  function run(){
    root.classList.add("is-in");
    bars.forEach(function(b){ b[0].style.width=pct(b[1])+"%"; });
    var reduced=window.matchMedia&&window.matchMedia("(prefers-reduced-motion: reduce)").matches;
    if(reduced){ vals.forEach(function(v){ v[0].textContent=v[1].toLocaleString(); }); return; }
    var t0=0;
    function tick(ts){
      if(!t0) t0=ts;
      var e=ts-t0, done=true;
      vals.forEach(function(v,i){
        var st=i*70, p=Math.min(1,Math.max(0,(e-st)/1000));
        if(p<1) done=false;
        var k=1-Math.pow(1-p,3);
        v[0].textContent=Math.round(MIN+(v[1]-MIN)*k).toLocaleString();
      });
      if(!done) requestAnimationFrame(tick);
    }
    requestAnimationFrame(tick);
  }

  var wasMob=null, raf;
  function relayout(){
    cancelAnimationFrame(raf);
    raf=requestAnimationFrame(function(){
      var m=mob();
      if(m!==wasMob){
        wasMob=m; build();
        if(root.classList.contains("is-in")){
          bars.forEach(function(b){ b[0].style.transitionDelay="0ms"; b[0].style.width=pct(b[1])+"%"; });
          vals.forEach(function(v){ v[0].textContent=v[1].toLocaleString(); });
        }
      } else {
        positionGrid(plot.querySelector(".el-grid"), m);
      }
    });
  }
  relayout();
  if(window.ResizeObserver){ new ResizeObserver(relayout).observe(plot); }
  else { window.addEventListener("resize",relayout); }

  if("IntersectionObserver" in window){
    var io=new IntersectionObserver(function(es){
      es.forEach(function(e){ if(e.isIntersecting){ io.unobserve(e.target); run(); } });
    },{threshold:0, rootMargin:"0px 0px -8% 0px"});
    io.observe(root);
  } else { run(); }
})();
</script>

Advancing across editing use cases

Great image editing comes down to control over the details that matter. Like moving a product into a new setting. Shifting the lighting and mood. Reworking individual objects without losing the bigger picture.

MAI-Image-2.7 improves on 2.6 across all 10 use cases and all 7 editing actions evaluated by Artificial Analysis. It reaches the category frontier in enhancement and restoration, reasoning-based edits, scene and style, and text or symbol editing. Its strongest lead is in scene and style, where it scores 24 Elo ahead of the next-best model.

<meta http-equiv="Content-Security-Policy" content="default-src 'none'; script-src 'unsafe-inline' 'unsafe-eval'; style-src 'unsafe-inline'; img-src data: blob: https://microsoft.ai/wp-content/uploads/; font-src data: https://microsoft.ai; media-src data: blob: https://microsoft.ai/wp-content/uploads/;"><div class="mai-climb" id="maiClimb">
  <header class="cb-hd">
    <h2 class="cb-title">Climb to the frontier</h2>
    <p class="cb-sub">Biggest category rank gains to #1, versus the previous generation</p>
  </header>

  <div class="cb-plot" id="cbPlot" role="img" aria-label="Category rank climbs from MAI-Image-2.6 to MAI-Image-2.7. Reasoning-based edit from rank 7 to rank 1, up 6 places. Scene and style edit from rank 7 to rank 1, up 6 places. Text or symbol edits from rank 6 to rank 1, up 5 places."></div>

  <ul class="cb-key" aria-hidden="true">
    <li><i class="cb-sw cb-sw-prev"></i>MAI-Image-2.6</li>
    <li><i class="cb-sw cb-sw-now"></i><em>MAI-Image-2.7 &middot; #1</em></li>
    <li class="cb-note">+ = places climbed</li>
  </ul>

  <p class="cb-src">Source: Artificial Analysis Image Arena &middot; Evaluated 22 Oct 2026</p>
</div>

<style>
@font-face{font-family:"Bradford LL";font-weight:450;font-style:normal;font-display:swap;src:url("https://microsoft.ai/wp-content/themes/wp-base-theme/dist/assets/BradfordLLWeb-Regular-C6SjiFc8.woff2") format("woff2");}@font-face{font-family:"Bradford LL";font-weight:450;font-style:italic;font-display:swap;src:url("https://microsoft.ai/wp-content/themes/wp-base-theme/dist/assets/BradfordLLWeb-Italic-CA7el69X.woff2") format("woff2");}@font-face{font-family:"Bradford LL";font-weight:600;font-style:normal;font-display:swap;src:url("https://microsoft.ai/wp-content/themes/wp-base-theme/dist/assets/BradfordLLWeb-Bold-BJqvqpMA.woff2") format("woff2");}@font-face{font-family:"Bradford LL";font-weight:600;font-style:italic;font-display:swap;src:url("https://microsoft.ai/wp-content/themes/wp-base-theme/dist/assets/BradfordLLWeb-BoldItalic-B6a2GBYT.woff2") format("woff2");}@font-face{font-family:"Red Hat Mono";font-weight:400;font-style:normal;font-display:swap;src:url("https://microsoft.ai/wp-content/themes/wp-base-theme/dist/assets/RedHatMono-Regular-CEF4ave-.woff2") format("woff2");}@font-face{font-family:"Red Hat Mono";font-weight:500;font-style:normal;font-display:swap;src:url("https://microsoft.ai/wp-content/themes/wp-base-theme/dist/assets/RedHatMono-Medium-BxJECwMl.woff2") format("woff2");}
html,body{margin:0;}
body{background:transparent;}
.mai-climb,.mai-climb *,.mai-climb *::before,.mai-climb *::after{box-sizing:border-box;}

.mai-climb{
  font-family:"Bradford LL", Georgia, serif;
  font-weight:400;
  color:#5D524B;
  -webkit-font-smoothing:antialiased;
}

.cb-hd{margin:0 0 40px;}
.cb-title{
  font-family:"Bradford LL", Georgia, serif;
  font-weight:400;font-size:32px;line-height:1.2;letter-spacing:-0.02em;margin:0;color:#5D524B;
}
.cb-sub{
  font-family:"Red Hat Mono",ui-monospace,Menlo,Consolas,monospace;
  font-weight:400;font-size:14px;line-height:1.5;margin:10px 0 0;color:#8A8077;
}

.cb-row{
  display:grid;
  grid-template-columns:230px minmax(0,1fr) 62px;
  align-items:center;
  column-gap:16px;
  margin-bottom:22px;
}
.cb-row:last-child{margin-bottom:0;}

.cb-name{
  font-family:"Bradford LL", Georgia, serif;
  font-weight:400;font-size:17px;line-height:1.25;text-align:right;color:#5D524B;
}

.cb-track{position:relative;height:34px;}
.cb-base{
  position:absolute;top:50%;left:0;right:0;height:2px;
  background:#E6DED0;transform:translateY(-50%);
}
.cb-seg{
  position:absolute;top:50%;height:3px;width:0;
  background:#738C7B;transform:translateY(-50%);
  transition:width 1s cubic-bezier(0.43,0.195,0.02,1);
}
.cb-dot{
  position:absolute;top:50%;
  border-radius:50%;display:flex;align-items:center;justify-content:center;
  font-family:"Red Hat Mono",ui-monospace,Menlo,Consolas,monospace;
  font-weight:500;font-variant-numeric:tabular-nums;
  transform:translate(-50%,-50%) scale(.6);opacity:0;
  transition:transform .55s cubic-bezier(0.43,0.195,0.02,1),
             opacity .45s cubic-bezier(0.43,0.195,0.02,1),
             left 1s cubic-bezier(0.43,0.195,0.02,1);
}
.is-in .cb-dot{transform:translate(-50%,-50%) scale(1);opacity:1;}
.cb-dot-prev{
  width:24px;height:24px;background:#D7DED6;color:#5D524B;font-size:11px;
}
.cb-dot-now{
  width:27px;height:27px;background:#738C7B;color:#FEF9ED;font-size:12px;
}

.cb-gain{
  font-family:"Bradford LL", Georgia, serif;
  font-weight:400;font-size:26px;line-height:1;color:#546B5C;
  text-align:right;font-variant-numeric:tabular-nums;
  opacity:0;transition:opacity .4s cubic-bezier(0.43,0.195,0.02,1);
}
.is-in .cb-gain{opacity:1;}

.cb-key{
  display:flex;flex-wrap:wrap;align-items:center;gap:26px;
  list-style:none;margin:30px 0 0;padding:0;
}
.cb-key li{
  display:inline-flex;align-items:center;gap:9px;
  font-family:"Bradford LL", Georgia, serif;
  font-weight:400;font-size:15px;line-height:1.3;color:#5D524B;
}
.cb-key em{font-style:italic;color:#546B5C;}
.cb-sw{display:block;width:13px;height:13px;flex:none;border-radius:50%;}
.cb-sw-prev{background:#D7DED6;}
.cb-sw-now{background:#738C7B;}
.cb-note{
  font-family:"Red Hat Mono",ui-monospace,Menlo,Consolas,monospace;
  font-size:12px;color:#8A8077;
}
.cb-src{
  font-family:"Red Hat Mono",ui-monospace,Menlo,Consolas,monospace;
  font-weight:400;font-size:12px;line-height:1.5;margin:22px 0 0;color:#8A8077;
}

/* ---- mobile ---- */
@media (max-width:600px){
  .cb-hd{margin-bottom:26px;}
  .cb-title{font-size:23px;}
  .cb-sub{font-size:12px;margin-top:8px;}
  .cb-row{
    grid-template-columns:minmax(0,1fr) 54px;
    column-gap:10px;row-gap:4px;margin-bottom:24px;
  }
  .cb-name{grid-column:1 / -1;text-align:left;font-size:16px;}
  .cb-track{height:30px;}
  .cb-gain{font-size:22px;}
  .cb-key{gap:14px 18px;margin-top:22px;}
  .cb-key li{font-size:13px;}
  .cb-note{font-size:11px;}
  .cb-src{font-size:11px;margin-top:18px;}
}

@media (prefers-reduced-motion:reduce){
  .cb-seg,.cb-dot,.cb-gain{transition:none !important;}
}
</style>

<script>
(function(){
  // [category, rank under 2.6, rank under 2.7]
  var D=[
    ["Reasoning-based edit",7,1],
    ["Scene & style edit",7,1],
    ["Text or symbol edits",6,1]
  ];
  var WORST=10;           // rank 10 sits at the left end, rank 1 at the right
  function pos(r){ return 8 + ((WORST-r)/(WORST-1))*84; }

  var root=document.getElementById("maiClimb"), plot=document.getElementById("cbPlot");
  if(!root||!plot) return;
  function el(t,c){ var n=document.createElement(t); if(c) n.className=c; return n; }
  var segs=[], nows=[];

  D.forEach(function(d,i){
    var a=pos(d[1]), b=pos(d[2]);
    var row=el("div","cb-row");

    var nm=el("div","cb-name"); nm.textContent=d[0];

    var tr=el("div","cb-track");
    tr.appendChild(el("div","cb-base"));
    var seg=el("div","cb-seg");
    seg.style.left=a+"%";
    seg.style.transitionDelay=(i*110)+"ms";
    tr.appendChild(seg);

    var dp=el("div","cb-dot cb-dot-prev"); dp.textContent=d[1];
    dp.style.left=a+"%"; dp.style.transitionDelay=(i*110)+"ms";
    var dn=el("div","cb-dot cb-dot-now"); dn.textContent=d[2];
    dn.style.left=a+"%"; dn.style.transitionDelay=(i*110+120)+"ms";
    tr.appendChild(dp); tr.appendChild(dn);

    var gn=el("div","cb-gain"); gn.textContent="+"+(d[1]-d[2]);
    gn.style.transitionDelay=(i*110+420)+"ms";

    row.appendChild(nm); row.appendChild(tr); row.appendChild(gn);
    plot.appendChild(row);
    segs.push([seg,Math.abs(b-a)]); nows.push([dn,b]);
  });

  function run(){
    root.classList.add("is-in");
    segs.forEach(function(s){ s[0].style.width=s[1]+"%"; });
    nows.forEach(function(n){ n[0].style.left=n[1]+"%"; });
  }

  if("IntersectionObserver" in window){
    var io=new IntersectionObserver(function(es){
      es.forEach(function(e){ if(e.isIntersecting){ io.unobserve(e.target); run(); } });
    },{threshold:0, rootMargin:"0px 0px -8% 0px"});
    io.observe(root);
  } else { run(); }
})();
</script>

For creators, that means more freedom to restyle a scene, fine-tune an asset, or bring a complex creative brief to life. It’s another step toward making powerful image editing more practical for everyday creative work.

Try it today

Try MAI Image 2.7’s image editing capabilities today on Arena, Artificial Analysis and MAI Playground. We’re starting with editing, with text-to-image generation coming soon.

Microsoft AI 发布 MAI Image 2.7 图像编辑模型，在 Arena 和 Artificial Analysis 取得前列排名

MAI-Cyber-1.1-Flash

MAI-Code-1.1-Flash

MAI-Voice-2.1

MAI-Transcribe-2

MAI-Thinking-1

MAI Image 2.7 debuts above GPT Image 2.5 Flare on Artificial Analysis and #3 on Arena

Advancing across editing use cases

Try it today

Related Stories

Introducing MAI-Image-2.5-Pro and MAI-Voice-2-Flash

Introducing MAI-Image-2:for limitless creativity

Introducing MAI-Image-1, debuting in the top 10 on LMArena

What's new

Microsoft Store

Education

Business

Developer & IT

Company