{"id":5096,"date":"2026-05-28T20:32:44","date_gmt":"2026-05-28T10:32:44","guid":{"rendered":"https:\/\/mokudigital.com\/space\/microsoft-just-open-sourced-a-text-to-image-model-that-runs-on-your-laptop-and-rivals-flux-2\/"},"modified":"2026-05-28T20:32:44","modified_gmt":"2026-05-28T10:32:44","slug":"microsoft-just-open-sourced-a-text-to-image-model-that-runs-on-your-laptop-and-rivals-flux-2","status":"publish","type":"post","link":"https:\/\/mokudigital.com\/space\/microsoft-just-open-sourced-a-text-to-image-model-that-runs-on-your-laptop-and-rivals-flux-2\/","title":{"rendered":"Microsoft Just Open-Sourced a Text-to-Image Model That Runs on Your Laptop and Rivals FLUX"},"content":{"rendered":"<h2>Microsoft Just Open-Sourced a Text-to-Image Model That Runs on Your Laptop and Rivals FLUX<\/h2>\n<p>Microsoft dropped Lens this week: a 3.8-billion-parameter text-to-image model that generates at up to 1440&#215;1440 resolution, produces output competitive with FLUX and Stable Diffusion 3, and ships with an MIT license. That last part is the real headline. You can download it right now from Hugging Face, run it locally, and build products on top of it without asking anyone for permission.<\/p>\n<h2>What Lens Actually Does<\/h2>\n<p>Lens is a diffusion model built on a 48-block MMDiT architecture. It uses FLUX.2 latents and a multi-layer GPT-OSS text encoder, which means it follows complex prompts well and handles multilingual input out of the box. The training corpus, Lens-800M, uses long GPT-4.1 captions for dense visual description, which is how it punches above its weight class at 3.8B parameters.<\/p>\n<p>The model supports nine aspect ratios from 1:2 to 2:1 and generates at 1024 or 1440 base resolution. In practice, that means you can prompt for a tall portrait, a wide landscape, or a square and it adapts without retraining or cropping.<\/p>\n<h2>The Turbo Variant: Four Steps, Sub-Second Generation<\/h2>\n<p>This is the part that changes workflows. Microsoft also released <strong>Lens-Turbo<\/strong>, a distilled variant that generates images in just 4 denoising steps with a guidance scale of 1.0 (no classifier-free guidance needed). On a modern GPU, that translates to image generation in well under a second. The standard Lens model uses 20 steps, which is still fast. But Turbo is the one you want if you are iterating on prompts or running batch generation.<\/p>\n<p>Three checkpoints are available on Hugging Face:<\/p>\n<ul>\n<li><strong>microsoft\/Lens<\/strong> \u2014 RL-tuned for visual quality, 20 steps, CFG 5.0<\/li>\n<li><strong>microsoft\/Lens-Turbo<\/strong> \u2014 Distilled for 4-step fast generation, CFG 1.0<\/li>\n<li><strong>microsoft\/Lens-Base<\/strong> \u2014 Supervised base model, 50 steps, CFG 5.0<\/li>\n<\/ul>\n<h2>What the Output Looks Like<\/h2>\n<p>The gallery Microsoft published is genuinely strong. A crystal dragon refracting aurora borealis light through transparent faceted crystal. A Remington typewriter with visible ink density variations on cream paper. Fish and chips with crispy beer batter texture you can almost feel. A koi pond shot from above with fallen cherry blossom petals on the surface. These are not generic AI images. The prompt adherence is tight, the textures are detailed, and the text rendering in particular is remarkably clean for a model this size.<\/p>\n<p>Text-in-image has been the graveyard of smaller diffusion models. Lens handles it better than most. The typewriter sample reads &#8220;Chapter One: The Beginning&#8221; with slightly misaligned Courier characters and visible ink density variation, exactly how a real typewriter prints. A flower shop sign reads &#8220;Fresh Flowers Daily&#8221; in flowing cursive with decorative swashes. A Route 66 gas station sign nails retro rounded sans-serif. This is not GPT-Image level text rendering, but it is far above what 4B-class models typically manage.<\/p>\n<h2>How to Run It Right Now<\/h2>\n<p>Installation takes about five minutes if you have a CUDA GPU:<\/p>\n<p><code>conda create -n lens python=3.12 -y<br \/>\nconda activate lens<br \/>\nuv pip install torch torchvision --index-url https:\/\/download.pytorch.org\/whl\/cu126<br \/>\nuv pip install -r requirements.txt<\/code><\/p>\n<p>Then generate an image with the Python API:<\/p>\n<p><code>from lens import LensPipeline<br \/>\nimport torch<br \/>\npipe = LensPipeline.from_pretrained(\"microsoft\/Lens\", torch_dtype=torch.bfloat16).to(\"cuda\")<br \/>\nimage = pipe(prompt=\"A cat holding a sign that says hello world\", base_resolution=1440, aspect_ratio=\"1:1\", num_inference_steps=20, guidance_scale=5.0).images[0]\nimage.save(\"output.png\")<\/code><\/p>\n<p>For the fast variant, swap to Lens-Turbo and drop to 4 steps:<\/p>\n<p><code>pipe = LensPipeline.from_pretrained(\"microsoft\/Lens-Turbo\", torch_dtype=torch.bfloat16).to(\"cuda\")<br \/>\nimage = pipe(prompt=\"your prompt\", base_resolution=1440, aspect_ratio=\"1:1\", num_inference_steps=4, guidance_scale=1.0).images[0]<\/code><\/p>\n<p>If you are GPU-constrained, use CPU offloading: replace <code>.to(\"cuda\")<\/code> with <code>pipe.enable_model_cpu_offload()<\/code>. On A100 or V100 GPUs, add the <code>--disable_mxfp4<\/code> flag to dequantize the text encoder.<\/p>\n<h2>What This Unlocks<\/h2>\n<p>MIT-licensed, open-weights, runs locally, rivals FLUX quality. That combination has not existed in a Microsoft model before. Here is what becomes possible now:<\/p>\n<ul>\n<li><strong>Product image generation without API costs.<\/strong> Every prompt is free. Every iteration is free. No per-image pricing, no rate limits, no usage tracking. You can generate 10,000 product shots and pay nothing beyond the electricity.<\/li>\n<li><strong>Privacy-first generation.<\/strong> Proprietary product photos, brand assets, and confidential creative briefs never leave your machine. For agencies and brands with strict IP policies, this removes the biggest objection to AI image generation.<\/li>\n<li><strong>Custom fine-tuning.<\/strong> MIT license means you can fine-tune Lens on your own product photography, brand style, or visual language and deploy the result commercially. No usage restrictions, no revenue caps, no enterprise agreement needed.<\/li>\n<li><strong>Embedded generation in apps.<\/strong> Because Lens is compact and Turbo is fast, you can embed image generation directly into tools, editors, and workflows. Think Canva-style generation built into your internal design tool, or a product page builder that generates hero shots on the fly.<\/li>\n<li><strong>Rapid iteration loops.<\/strong> Lens-Turbo at 4 steps means you can generate, evaluate, tweak the prompt, and regenerate in the time it takes most cloud APIs to return a single image. The creative loop shrinks from minutes to seconds.<\/li>\n<\/ul>\n<h2>What to Look Out For<\/h2>\n<p>Lens is released for research purposes, and Microsoft notes it may generate inaccurate, biased, or inappropriate outputs under certain prompts. The training data comes from web-scale collections with known bias issues. If you are deploying it in a product, you need your own content moderation and safety layer on top.<\/p>\n<p>The model also requires a decent GPU for comfortable inference. CPU offloading works but slows things down. If you are on an older machine without a dedicated GPU, this is not the model for you (though PrismML&#8217;s Bonsai Image 4B, released the same week, targets exactly that use case with 1-bit quantization for iPhones and laptops).<\/p>\n<p>Multilingual prompt support via the GPT-OSS encoder is a real advantage over models that only work well in English. If you create content in multiple languages, Lens should handle non-English prompts more reliably than most open-source alternatives.<\/p>\n<h2>The Bigger Picture<\/h2>\n<p>Two things happened this week that move image generation closer to the user. Microsoft open-sourced Lens with MIT licensing and sub-second Turbo inference. PrismML released Bonsai Image 4B, a 1-bit quantized model that runs on an iPhone. Together, they signal that the &#8220;cloud API or nothing&#8221; era of image generation is ending. The models are getting small enough, fast enough, and good enough to live on your hardware. The economics flip when generation is free, private, and instant.<\/p>\n<p>Download Lens from <a href=\"https:\/\/huggingface.co\/microsoft\/Lens\">Hugging Face<\/a>. Grab the code from <a href=\"https:\/\/github.com\/microsoft\/Lens\">GitHub<\/a>. Read the <a href=\"https:\/\/arxiv.org\/abs\/2605.21573\">paper on arXiv<\/a>. Start generating.<\/p>\n<p><em>Want more like this? Join MOKU Club for free. Weekly resources, early access to new guides, and occasional templates you can actually use. Join below.<\/em><\/p>\n","protected":false},"excerpt":{"rendered":"<p>Microsoft Just Open-Sourced a Text-to-Image Model That Runs on Your Laptop and Rivals FLUX Microsoft dropped Lens this week: a 3.8-billion-parameter text-to-image model that generates at up to 1440&#215;1440 resolution,&#8230;<\/p>\n","protected":false},"author":8,"featured_media":5094,"comment_status":"closed","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"inline_featured_image":false,"footnotes":""},"categories":[71,64,60],"tags":[],"class_list":["post-5096","post","type-post","status-publish","format-standard","has-post-thumbnail","category-ai","category-design","category-news"],"yoast_head":"<!-- This site is optimized with the Yoast SEO plugin v27.4 - https:\/\/yoast.com\/product\/yoast-seo-wordpress\/ -->\n<title>Microsoft Just Open-Sourced a Text-to-Image Model That Runs on Your Laptop and Rivals FLUX &#8212; MOKU Space<\/title>\n<meta name=\"robots\" content=\"index, follow, max-snippet:-1, max-image-preview:large, max-video-preview:-1\" \/>\n<link rel=\"canonical\" href=\"https:\/\/mokudigital.com\/space\/microsoft-just-open-sourced-a-text-to-image-model-that-runs-on-your-laptop-and-rivals-flux-2\/\" \/>\n<meta property=\"og:locale\" content=\"en_US\" \/>\n<meta property=\"og:type\" content=\"article\" \/>\n<meta property=\"og:title\" content=\"Microsoft Just Open-Sourced a Text-to-Image Model That Runs on Your Laptop and Rivals FLUX &#8212; MOKU Space\" \/>\n<meta property=\"og:description\" content=\"Microsoft Just Open-Sourced a Text-to-Image Model That Runs on Your Laptop and Rivals FLUX Microsoft dropped Lens this week: a 3.8-billion-parameter text-to-image model that generates at up to 1440&#215;1440 resolution,...\" \/>\n<meta property=\"og:url\" content=\"https:\/\/mokudigital.com\/space\/microsoft-just-open-sourced-a-text-to-image-model-that-runs-on-your-laptop-and-rivals-flux-2\/\" \/>\n<meta property=\"og:site_name\" content=\"MOKU Space\" \/>\n<meta property=\"article:publisher\" content=\"https:\/\/www.facebook.com\/MOKUdigital\/\" \/>\n<meta property=\"article:published_time\" content=\"2026-05-28T10:32:44+00:00\" \/>\n<meta property=\"og:image\" content=\"https:\/\/mokudigital.com\/space\/wp-content\/uploads\/2024\/02\/mokuspace_og.jpg\" \/>\n\t<meta property=\"og:image:width\" content=\"1200\" \/>\n\t<meta property=\"og:image:height\" content=\"675\" \/>\n\t<meta property=\"og:image:type\" content=\"image\/jpeg\" \/>\n<meta name=\"author\" content=\"Hera\" \/>\n<meta name=\"twitter:card\" content=\"summary_large_image\" \/>\n<meta name=\"twitter:label1\" content=\"Written by\" \/>\n\t<meta name=\"twitter:data1\" content=\"Hera\" \/>\n\t<meta name=\"twitter:label2\" content=\"Est. reading time\" \/>\n\t<meta name=\"twitter:data2\" content=\"5 minutes\" \/>\n<script type=\"application\/ld+json\" class=\"yoast-schema-graph\">{\"@context\":\"https:\\\/\\\/schema.org\",\"@graph\":[{\"@type\":\"Article\",\"@id\":\"https:\\\/\\\/mokudigital.com\\\/space\\\/microsoft-just-open-sourced-a-text-to-image-model-that-runs-on-your-laptop-and-rivals-flux-2\\\/#article\",\"isPartOf\":{\"@id\":\"https:\\\/\\\/mokudigital.com\\\/space\\\/microsoft-just-open-sourced-a-text-to-image-model-that-runs-on-your-laptop-and-rivals-flux-2\\\/\"},\"author\":{\"name\":\"Hera\",\"@id\":\"https:\\\/\\\/mokudigital.com\\\/space\\\/#\\\/schema\\\/person\\\/fcc0cf0192c88b70fc3156df50cbde7d\"},\"headline\":\"Microsoft Just Open-Sourced a Text-to-Image Model That Runs on Your Laptop and Rivals FLUX\",\"datePublished\":\"2026-05-28T10:32:44+00:00\",\"mainEntityOfPage\":{\"@id\":\"https:\\\/\\\/mokudigital.com\\\/space\\\/microsoft-just-open-sourced-a-text-to-image-model-that-runs-on-your-laptop-and-rivals-flux-2\\\/\"},\"wordCount\":1005,\"publisher\":{\"@id\":\"https:\\\/\\\/mokudigital.com\\\/space\\\/#organization\"},\"image\":{\"@id\":\"https:\\\/\\\/mokudigital.com\\\/space\\\/microsoft-just-open-sourced-a-text-to-image-model-that-runs-on-your-laptop-and-rivals-flux-2\\\/#primaryimage\"},\"thumbnailUrl\":\"https:\\\/\\\/mokudigital.com\\\/space\\\/wp-content\\\/uploads\\\/2026\\\/05\\\/hero-lens-open-source-0528-2.webp\",\"articleSection\":[\"Ai\",\"Design\",\"News\"],\"inLanguage\":\"en-US\"},{\"@type\":\"WebPage\",\"@id\":\"https:\\\/\\\/mokudigital.com\\\/space\\\/microsoft-just-open-sourced-a-text-to-image-model-that-runs-on-your-laptop-and-rivals-flux-2\\\/\",\"url\":\"https:\\\/\\\/mokudigital.com\\\/space\\\/microsoft-just-open-sourced-a-text-to-image-model-that-runs-on-your-laptop-and-rivals-flux-2\\\/\",\"name\":\"Microsoft Just Open-Sourced a Text-to-Image Model That Runs on Your Laptop and Rivals FLUX &#8212; MOKU Space\",\"isPartOf\":{\"@id\":\"https:\\\/\\\/mokudigital.com\\\/space\\\/#website\"},\"primaryImageOfPage\":{\"@id\":\"https:\\\/\\\/mokudigital.com\\\/space\\\/microsoft-just-open-sourced-a-text-to-image-model-that-runs-on-your-laptop-and-rivals-flux-2\\\/#primaryimage\"},\"image\":{\"@id\":\"https:\\\/\\\/mokudigital.com\\\/space\\\/microsoft-just-open-sourced-a-text-to-image-model-that-runs-on-your-laptop-and-rivals-flux-2\\\/#primaryimage\"},\"thumbnailUrl\":\"https:\\\/\\\/mokudigital.com\\\/space\\\/wp-content\\\/uploads\\\/2026\\\/05\\\/hero-lens-open-source-0528-2.webp\",\"datePublished\":\"2026-05-28T10:32:44+00:00\",\"breadcrumb\":{\"@id\":\"https:\\\/\\\/mokudigital.com\\\/space\\\/microsoft-just-open-sourced-a-text-to-image-model-that-runs-on-your-laptop-and-rivals-flux-2\\\/#breadcrumb\"},\"inLanguage\":\"en-US\",\"potentialAction\":[{\"@type\":\"ReadAction\",\"target\":[\"https:\\\/\\\/mokudigital.com\\\/space\\\/microsoft-just-open-sourced-a-text-to-image-model-that-runs-on-your-laptop-and-rivals-flux-2\\\/\"]}]},{\"@type\":\"ImageObject\",\"inLanguage\":\"en-US\",\"@id\":\"https:\\\/\\\/mokudigital.com\\\/space\\\/microsoft-just-open-sourced-a-text-to-image-model-that-runs-on-your-laptop-and-rivals-flux-2\\\/#primaryimage\",\"url\":\"https:\\\/\\\/mokudigital.com\\\/space\\\/wp-content\\\/uploads\\\/2026\\\/05\\\/hero-lens-open-source-0528-2.webp\",\"contentUrl\":\"https:\\\/\\\/mokudigital.com\\\/space\\\/wp-content\\\/uploads\\\/2026\\\/05\\\/hero-lens-open-source-0528-2.webp\",\"width\":1024,\"height\":1536},{\"@type\":\"BreadcrumbList\",\"@id\":\"https:\\\/\\\/mokudigital.com\\\/space\\\/microsoft-just-open-sourced-a-text-to-image-model-that-runs-on-your-laptop-and-rivals-flux-2\\\/#breadcrumb\",\"itemListElement\":[{\"@type\":\"ListItem\",\"position\":1,\"name\":\"Home\",\"item\":\"https:\\\/\\\/mokudigital.com\\\/space\\\/\"},{\"@type\":\"ListItem\",\"position\":2,\"name\":\"Microsoft Just Open-Sourced a Text-to-Image Model That Runs on Your Laptop and Rivals FLUX\"}]},{\"@type\":\"WebSite\",\"@id\":\"https:\\\/\\\/mokudigital.com\\\/space\\\/#website\",\"url\":\"https:\\\/\\\/mokudigital.com\\\/space\\\/\",\"name\":\"MOKU Space\",\"description\":\"A space to learn and find resources for growth in the digital landscape.\",\"publisher\":{\"@id\":\"https:\\\/\\\/mokudigital.com\\\/space\\\/#organization\"},\"potentialAction\":[{\"@type\":\"SearchAction\",\"target\":{\"@type\":\"EntryPoint\",\"urlTemplate\":\"https:\\\/\\\/mokudigital.com\\\/space\\\/?s={search_term_string}\"},\"query-input\":{\"@type\":\"PropertyValueSpecification\",\"valueRequired\":true,\"valueName\":\"search_term_string\"}}],\"inLanguage\":\"en-US\"},{\"@type\":\"Organization\",\"@id\":\"https:\\\/\\\/mokudigital.com\\\/space\\\/#organization\",\"name\":\"MOKU Digital\",\"alternateName\":\"MOKU\",\"url\":\"https:\\\/\\\/mokudigital.com\\\/space\\\/\",\"logo\":{\"@type\":\"ImageObject\",\"inLanguage\":\"en-US\",\"@id\":\"https:\\\/\\\/mokudigital.com\\\/space\\\/#\\\/schema\\\/logo\\\/image\\\/\",\"url\":\"https:\\\/\\\/mokudigital.com\\\/space\\\/wp-content\\\/uploads\\\/2024\\\/01\\\/MOKU_SPACE_ICON_2x.png\",\"contentUrl\":\"https:\\\/\\\/mokudigital.com\\\/space\\\/wp-content\\\/uploads\\\/2024\\\/01\\\/MOKU_SPACE_ICON_2x.png\",\"width\":580,\"height\":579,\"caption\":\"MOKU Digital\"},\"image\":{\"@id\":\"https:\\\/\\\/mokudigital.com\\\/space\\\/#\\\/schema\\\/logo\\\/image\\\/\"},\"sameAs\":[\"https:\\\/\\\/www.facebook.com\\\/MOKUdigital\\\/\",\"https:\\\/\\\/www.instagram.com\\\/mokudigital\"]},{\"@type\":\"Person\",\"@id\":\"https:\\\/\\\/mokudigital.com\\\/space\\\/#\\\/schema\\\/person\\\/fcc0cf0192c88b70fc3156df50cbde7d\",\"name\":\"Hera\",\"image\":{\"@type\":\"ImageObject\",\"inLanguage\":\"en-US\",\"@id\":\"https:\\\/\\\/mokudigital.com\\\/space\\\/wp-content\\\/uploads\\\/2026\\\/04\\\/cropped-hera_pfp-bw_v2-96x96.jpg\",\"url\":\"https:\\\/\\\/mokudigital.com\\\/space\\\/wp-content\\\/uploads\\\/2026\\\/04\\\/cropped-hera_pfp-bw_v2-96x96.jpg\",\"contentUrl\":\"https:\\\/\\\/mokudigital.com\\\/space\\\/wp-content\\\/uploads\\\/2026\\\/04\\\/cropped-hera_pfp-bw_v2-96x96.jpg\",\"caption\":\"Hera\"},\"url\":\"https:\\\/\\\/mokudigital.com\\\/space\\\/author\\\/hera\\\/\"}]}<\/script>\n<!-- \/ Yoast SEO plugin. -->","yoast_head_json":{"title":"Microsoft Just Open-Sourced a Text-to-Image Model That Runs on Your Laptop and Rivals FLUX &#8212; MOKU Space","robots":{"index":"index","follow":"follow","max-snippet":"max-snippet:-1","max-image-preview":"max-image-preview:large","max-video-preview":"max-video-preview:-1"},"canonical":"https:\/\/mokudigital.com\/space\/microsoft-just-open-sourced-a-text-to-image-model-that-runs-on-your-laptop-and-rivals-flux-2\/","og_locale":"en_US","og_type":"article","og_title":"Microsoft Just Open-Sourced a Text-to-Image Model That Runs on Your Laptop and Rivals FLUX &#8212; MOKU Space","og_description":"Microsoft Just Open-Sourced a Text-to-Image Model That Runs on Your Laptop and Rivals FLUX Microsoft dropped Lens this week: a 3.8-billion-parameter text-to-image model that generates at up to 1440&#215;1440 resolution,...","og_url":"https:\/\/mokudigital.com\/space\/microsoft-just-open-sourced-a-text-to-image-model-that-runs-on-your-laptop-and-rivals-flux-2\/","og_site_name":"MOKU Space","article_publisher":"https:\/\/www.facebook.com\/MOKUdigital\/","article_published_time":"2026-05-28T10:32:44+00:00","og_image":[{"width":1200,"height":675,"url":"https:\/\/mokudigital.com\/space\/wp-content\/uploads\/2024\/02\/mokuspace_og.jpg","type":"image\/jpeg"}],"author":"Hera","twitter_card":"summary_large_image","twitter_misc":{"Written by":"Hera","Est. reading time":"5 minutes"},"schema":{"@context":"https:\/\/schema.org","@graph":[{"@type":"Article","@id":"https:\/\/mokudigital.com\/space\/microsoft-just-open-sourced-a-text-to-image-model-that-runs-on-your-laptop-and-rivals-flux-2\/#article","isPartOf":{"@id":"https:\/\/mokudigital.com\/space\/microsoft-just-open-sourced-a-text-to-image-model-that-runs-on-your-laptop-and-rivals-flux-2\/"},"author":{"name":"Hera","@id":"https:\/\/mokudigital.com\/space\/#\/schema\/person\/fcc0cf0192c88b70fc3156df50cbde7d"},"headline":"Microsoft Just Open-Sourced a Text-to-Image Model That Runs on Your Laptop and Rivals FLUX","datePublished":"2026-05-28T10:32:44+00:00","mainEntityOfPage":{"@id":"https:\/\/mokudigital.com\/space\/microsoft-just-open-sourced-a-text-to-image-model-that-runs-on-your-laptop-and-rivals-flux-2\/"},"wordCount":1005,"publisher":{"@id":"https:\/\/mokudigital.com\/space\/#organization"},"image":{"@id":"https:\/\/mokudigital.com\/space\/microsoft-just-open-sourced-a-text-to-image-model-that-runs-on-your-laptop-and-rivals-flux-2\/#primaryimage"},"thumbnailUrl":"https:\/\/mokudigital.com\/space\/wp-content\/uploads\/2026\/05\/hero-lens-open-source-0528-2.webp","articleSection":["Ai","Design","News"],"inLanguage":"en-US"},{"@type":"WebPage","@id":"https:\/\/mokudigital.com\/space\/microsoft-just-open-sourced-a-text-to-image-model-that-runs-on-your-laptop-and-rivals-flux-2\/","url":"https:\/\/mokudigital.com\/space\/microsoft-just-open-sourced-a-text-to-image-model-that-runs-on-your-laptop-and-rivals-flux-2\/","name":"Microsoft Just Open-Sourced a Text-to-Image Model That Runs on Your Laptop and Rivals FLUX &#8212; MOKU Space","isPartOf":{"@id":"https:\/\/mokudigital.com\/space\/#website"},"primaryImageOfPage":{"@id":"https:\/\/mokudigital.com\/space\/microsoft-just-open-sourced-a-text-to-image-model-that-runs-on-your-laptop-and-rivals-flux-2\/#primaryimage"},"image":{"@id":"https:\/\/mokudigital.com\/space\/microsoft-just-open-sourced-a-text-to-image-model-that-runs-on-your-laptop-and-rivals-flux-2\/#primaryimage"},"thumbnailUrl":"https:\/\/mokudigital.com\/space\/wp-content\/uploads\/2026\/05\/hero-lens-open-source-0528-2.webp","datePublished":"2026-05-28T10:32:44+00:00","breadcrumb":{"@id":"https:\/\/mokudigital.com\/space\/microsoft-just-open-sourced-a-text-to-image-model-that-runs-on-your-laptop-and-rivals-flux-2\/#breadcrumb"},"inLanguage":"en-US","potentialAction":[{"@type":"ReadAction","target":["https:\/\/mokudigital.com\/space\/microsoft-just-open-sourced-a-text-to-image-model-that-runs-on-your-laptop-and-rivals-flux-2\/"]}]},{"@type":"ImageObject","inLanguage":"en-US","@id":"https:\/\/mokudigital.com\/space\/microsoft-just-open-sourced-a-text-to-image-model-that-runs-on-your-laptop-and-rivals-flux-2\/#primaryimage","url":"https:\/\/mokudigital.com\/space\/wp-content\/uploads\/2026\/05\/hero-lens-open-source-0528-2.webp","contentUrl":"https:\/\/mokudigital.com\/space\/wp-content\/uploads\/2026\/05\/hero-lens-open-source-0528-2.webp","width":1024,"height":1536},{"@type":"BreadcrumbList","@id":"https:\/\/mokudigital.com\/space\/microsoft-just-open-sourced-a-text-to-image-model-that-runs-on-your-laptop-and-rivals-flux-2\/#breadcrumb","itemListElement":[{"@type":"ListItem","position":1,"name":"Home","item":"https:\/\/mokudigital.com\/space\/"},{"@type":"ListItem","position":2,"name":"Microsoft Just Open-Sourced a Text-to-Image Model That Runs on Your Laptop and Rivals FLUX"}]},{"@type":"WebSite","@id":"https:\/\/mokudigital.com\/space\/#website","url":"https:\/\/mokudigital.com\/space\/","name":"MOKU Space","description":"A space to learn and find resources for growth in the digital landscape.","publisher":{"@id":"https:\/\/mokudigital.com\/space\/#organization"},"potentialAction":[{"@type":"SearchAction","target":{"@type":"EntryPoint","urlTemplate":"https:\/\/mokudigital.com\/space\/?s={search_term_string}"},"query-input":{"@type":"PropertyValueSpecification","valueRequired":true,"valueName":"search_term_string"}}],"inLanguage":"en-US"},{"@type":"Organization","@id":"https:\/\/mokudigital.com\/space\/#organization","name":"MOKU Digital","alternateName":"MOKU","url":"https:\/\/mokudigital.com\/space\/","logo":{"@type":"ImageObject","inLanguage":"en-US","@id":"https:\/\/mokudigital.com\/space\/#\/schema\/logo\/image\/","url":"https:\/\/mokudigital.com\/space\/wp-content\/uploads\/2024\/01\/MOKU_SPACE_ICON_2x.png","contentUrl":"https:\/\/mokudigital.com\/space\/wp-content\/uploads\/2024\/01\/MOKU_SPACE_ICON_2x.png","width":580,"height":579,"caption":"MOKU Digital"},"image":{"@id":"https:\/\/mokudigital.com\/space\/#\/schema\/logo\/image\/"},"sameAs":["https:\/\/www.facebook.com\/MOKUdigital\/","https:\/\/www.instagram.com\/mokudigital"]},{"@type":"Person","@id":"https:\/\/mokudigital.com\/space\/#\/schema\/person\/fcc0cf0192c88b70fc3156df50cbde7d","name":"Hera","image":{"@type":"ImageObject","inLanguage":"en-US","@id":"https:\/\/mokudigital.com\/space\/wp-content\/uploads\/2026\/04\/cropped-hera_pfp-bw_v2-96x96.jpg","url":"https:\/\/mokudigital.com\/space\/wp-content\/uploads\/2026\/04\/cropped-hera_pfp-bw_v2-96x96.jpg","contentUrl":"https:\/\/mokudigital.com\/space\/wp-content\/uploads\/2026\/04\/cropped-hera_pfp-bw_v2-96x96.jpg","caption":"Hera"},"url":"https:\/\/mokudigital.com\/space\/author\/hera\/"}]}},"_links":{"self":[{"href":"https:\/\/mokudigital.com\/space\/wp-json\/wp\/v2\/posts\/5096","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/mokudigital.com\/space\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/mokudigital.com\/space\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/mokudigital.com\/space\/wp-json\/wp\/v2\/users\/8"}],"replies":[{"embeddable":true,"href":"https:\/\/mokudigital.com\/space\/wp-json\/wp\/v2\/comments?post=5096"}],"version-history":[{"count":0,"href":"https:\/\/mokudigital.com\/space\/wp-json\/wp\/v2\/posts\/5096\/revisions"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/mokudigital.com\/space\/wp-json\/wp\/v2\/media\/5094"}],"wp:attachment":[{"href":"https:\/\/mokudigital.com\/space\/wp-json\/wp\/v2\/media?parent=5096"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/mokudigital.com\/space\/wp-json\/wp\/v2\/categories?post=5096"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/mokudigital.com\/space\/wp-json\/wp\/v2\/tags?post=5096"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}