{"id":71,"date":"2026-05-18T12:37:31","date_gmt":"2026-05-18T12:37:31","guid":{"rendered":"https:\/\/convly.ai\/image-generation-models-comparison\/"},"modified":"2026-06-10T05:06:01","modified_gmt":"2026-06-10T05:06:01","slug":"image-generation-models-comparison","status":"publish","type":"post","link":"https:\/\/convly.ai\/pt\/image-generation-models-comparison\/","title":{"rendered":"Modelos de gera\u00e7\u00e3o de imagens por IA em 2026: como funcionam e quais usar"},"content":{"rendered":"<p>Most &#8220;AI image generator&#8221; comparisons rank apps. This one goes a layer deeper, to the <strong>modelos<\/strong> those apps are built on \u2014 because if you&#8217;re a developer, a power user, or someone choosing what to build a product on, the model is what actually matters. The same model can power three different apps; understanding the model tells you what&#8217;s really possible.<\/p>\n<p>This guide explains how 2026&#8217;s image generation models work and compares the major model families on the things that matter when you pick one to build with.<\/p>\n<div class=\"convly-tldr\">\n<h3>Principais conclus\u00f5es<\/h3>\n<ul>\n<li><strong>Two architectures dominate:<\/strong> diffusion models (most generators) and autoregressive\/transformer models (GPT-4o-style native image generation).<\/li>\n<li><strong>Best open model:<\/strong> FLUX \u2014 the de facto standard for self-hosted, customizable image generation.<\/li>\n<li><strong>Best for prompt precision:<\/strong> autoregressive models like GPT-4o&#8217;s native image generation.<\/li>\n<li><strong>Best for fine-tuning:<\/strong> the Stable Diffusion \/ FLUX open ecosystem, with LoRAs and full control.<\/li>\n<li><strong>Closed models<\/strong> (Midjourney&#8217;s, Imagen) lead on polish but can&#8217;t be self-hosted or deeply customized.<\/li>\n<\/ul>\n<\/div>\n<div id=\"ez-toc-container\" class=\"ez-toc-v2_0_84 counter-flat ez-toc-counter ez-toc-container-direction\">\n<label for=\"ez-toc-cssicon-toggle-item-6a38b815aaf38\" class=\"ez-toc-cssicon-toggle-label\"><span class=\"\"><span class=\"eztoc-hide\" style=\"display:none;\">Alternar<\/span><span class=\"ez-toc-icon-toggle-span\"><svg style=\"fill: #000000;color:#000000\" xmlns=\"http:\/\/www.w3.org\/2000\/svg\" class=\"list-377408\" width=\"20px\" height=\"20px\" viewbox=\"0 0 24 24\" fill=\"none\"><path d=\"M6 6H4v2h2V6zm14 0H8v2h12V6zM4 11h2v2H4v-2zm16 0H8v2h12v-2zM4 16h2v2H4v-2zm16 0H8v2h12v-2z\" fill=\"currentColor\"><\/path><\/svg><svg style=\"fill: #000000;color:#000000\" class=\"arrow-unsorted-368013\" xmlns=\"http:\/\/www.w3.org\/2000\/svg\" width=\"10px\" height=\"10px\" viewbox=\"0 0 24 24\" version=\"1.2\" baseprofile=\"tiny\"><path d=\"M18.2 9.3l-6.2-6.3-6.2 6.3c-.2.2-.3.4-.3.7s.1.5.3.7c.2.2.4.3.7.3h11c.3 0 .5-.1.7-.3.2-.2.3-.5.3-.7s-.1-.5-.3-.7zM5.8 14.7l6.2 6.3 6.2-6.3c.2-.2.3-.5.3-.7s-.1-.5-.3-.7c-.2-.2-.4-.3-.7-.3h-11c-.3 0-.5.1-.7.3-.2.2-.3.5-.3.7s.1.5.3.7z\"\/><\/svg><\/span><\/span><\/label><input type=\"checkbox\"  id=\"ez-toc-cssicon-toggle-item-6a38b815aaf38\"  aria-label=\"Alternar\" \/><nav><ul class='ez-toc-list ez-toc-list-level-1' ><li class='ez-toc-page-1'><a class=\"ez-toc-link ez-toc-heading-1\" href=\"https:\/\/convly.ai\/pt\/image-generation-models-comparison\/#How_AI_image_models_work\" >How AI image models work<\/a><\/li><li class='ez-toc-page-1'><a class=\"ez-toc-link ez-toc-heading-2\" href=\"https:\/\/convly.ai\/pt\/image-generation-models-comparison\/#The_major_model_families\" >The major model families<\/a><\/li><li class='ez-toc-page-1'><a class=\"ez-toc-link ez-toc-heading-3\" href=\"https:\/\/convly.ai\/pt\/image-generation-models-comparison\/#Side-by-side_comparison\" >Side-by-side comparison<\/a><\/li><li class='ez-toc-page-1'><a class=\"ez-toc-link ez-toc-heading-4\" href=\"https:\/\/convly.ai\/pt\/image-generation-models-comparison\/#Which_model_should_you_build_on\" >Which model should you build on?<\/a><\/li><li class='ez-toc-page-1'><a class=\"ez-toc-link ez-toc-heading-5\" href=\"https:\/\/convly.ai\/pt\/image-generation-models-comparison\/#Open_vs_closed_the_real_trade-off\" >Open vs closed: the real trade-off<\/a><\/li><li class='ez-toc-page-1'><a class=\"ez-toc-link ez-toc-heading-6\" href=\"https:\/\/convly.ai\/pt\/image-generation-models-comparison\/#What_it_costs_to_generate_images_at_scale\" >What it costs to generate images at scale<\/a><\/li><li class='ez-toc-page-1'><a class=\"ez-toc-link ez-toc-heading-7\" href=\"https:\/\/convly.ai\/pt\/image-generation-models-comparison\/#FAQ\" >Perguntas frequentes<\/a><\/li><li class='ez-toc-page-1'><a class=\"ez-toc-link ez-toc-heading-8\" href=\"https:\/\/convly.ai\/pt\/image-generation-models-comparison\/#Bottom_line\" >Conclus\u00e3o<\/a><\/li><li class='ez-toc-page-1'><a class=\"ez-toc-link ez-toc-heading-9\" href=\"https:\/\/convly.ai\/pt\/image-generation-models-comparison\/#Related_articles\" >Artigos relacionados<\/a><\/li><\/ul><\/nav><\/div>\n<h2><span class=\"ez-toc-section\" id=\"How_AI_image_models_work\"><\/span>How AI image models work<span class=\"ez-toc-section-end\"><\/span><\/h2>\n<p>Two architectures power almost everything in 2026.<\/p>\n<h3>Diffusion models<\/h3>\n<p>Diffusion is the technique behind Stable Diffusion, FLUX, Midjourney, Imagen, and most generators. The idea: take a training image, add noise step by step until it&#8217;s pure static, then train a model to <em>reverse<\/em> that process. To generate a new image, the model starts from random noise and progressively &#8220;denoises&#8221; it into a coherent picture, guided by your text prompt.<\/p>\n<p>Diffusion models are excellent at texture, lighting, and overall image quality. Their classic weakness is precise control \u2014 counting objects, placing them exactly, rendering specific text \u2014 because they shape the whole image at once rather than reasoning about it part by part.<\/p>\n<h3>Autoregressive (transformer) models<\/h3>\n<p>The newer approach, used by GPT-4o&#8217;s native image generation, treats an image more like language: the model generates it as a sequence, predicting image tokens in order, the same way a language model predicts words.<\/p>\n<p>Because this approach shares architecture with large language models, it inherits their strength: <strong>understanding<\/strong>. Autoregressive image models follow complex instructions, render text, and respect spatial relationships better than pure diffusion. The trade-off is that generation can be slower and, historically, slightly less painterly \u2014 though that gap has largely closed.<\/p>\n<p>Many 2026 systems are effectively hybrids, combining the instruction-following of transformers with the visual quality of diffusion.<\/p>\n<h2><span class=\"ez-toc-section\" id=\"The_major_model_families\"><\/span>The major model families<span class=\"ez-toc-section-end\"><\/span><\/h2>\n<h3>FLUX (Black Forest Labs)<\/h3>\n<p>FLUX is the open-weight leader in 2026. It offers excellent quality, strong prompt adherence, and decent text rendering \u2014 and it&#8217;s available as downloadable weights you can run, fine-tune, and embed in products. It comes in variants tuned for speed versus maximum quality. For most builders who want an open model, FLUX is the default starting point.<\/p>\n<h3>Stable Diffusion (3.5 line)<\/h3>\n<p>Stable Diffusion is the model family that created the open image-AI ecosystem. The 3.5-generation models remain widely used, and the surrounding tooling \u2014 fine-tuning pipelines, LoRAs, ControlNet-style guidance, a huge library of community checkpoints \u2014 is unmatched. If you need deep customization and a mature toolchain, the Stable Diffusion ecosystem is still the richest, even as FLUX leads on raw quality.<\/p>\n<h3>GPT-4o native image generation (OpenAI)<\/h3>\n<p>OpenAI&#8217;s autoregressive image model is the benchmark for prompt precision and conversational editing. It&#8217;s closed and API-only \u2014 you can&#8217;t self-host it \u2014 but for applications that need an image to match a detailed brief, or to be edited through natural language, it&#8217;s the strongest option. Access is through OpenAI&#8217;s API.<\/p>\n<h3>Imagen (Google)<\/h3>\n<p>Imagen powers image generation in Gemini and Google&#8217;s creative tools. It&#8217;s a closed model with excellent photorealism and strong safety filtering, available through Google&#8217;s API. A solid choice if your stack is already on Google Cloud.<\/p>\n<h3>Midjourney&#8217;s model<\/h3>\n<p>Midjourney runs its own proprietary, closed model \u2014 the source of its signature aesthetic. It&#8217;s available only through Midjourney&#8217;s own app, with no API or self-hosting. You use it for the output; you can&#8217;t build on the model directly.<\/p>\n<h2><span class=\"ez-toc-section\" id=\"Side-by-side_comparison\"><\/span>Side-by-side comparison<span class=\"ez-toc-section-end\"><\/span><\/h2>\n<table class=\"convly-vs\">\n<thead>\n<tr>\n<th>Modelo<\/th>\n<th>Tipo<\/th>\n<th>Pesos abertos<\/th>\n<th>Strength<\/th>\n<th>Access<\/th>\n<\/tr>\n<\/thead>\n<tbody>\n<tr>\n<td>FLUX<\/td>\n<td>Diffusion<\/td>\n<td>Sim<\/td>\n<td>Open quality + customization<\/td>\n<td>Self-host or API<\/td>\n<\/tr>\n<tr>\n<td>Stable Diffusion 3.5<\/td>\n<td>Diffusion<\/td>\n<td>Sim<\/td>\n<td>Fine-tuning ecosystem<\/td>\n<td>Self-host or API<\/td>\n<\/tr>\n<tr>\n<td>GPT-4o image gen<\/td>\n<td>Autoregressive<\/td>\n<td>N\u00e3o<\/td>\n<td>Prompt precision, editing<\/td>\n<td>OpenAI API<\/td>\n<\/tr>\n<tr>\n<td>Imagen<\/td>\n<td>Diffusion<\/td>\n<td>N\u00e3o<\/td>\n<td>Fotorrealismo<\/td>\n<td>Google API<\/td>\n<\/tr>\n<tr>\n<td>Midjourney model<\/td>\n<td>Diffusion<\/td>\n<td>N\u00e3o<\/td>\n<td>Aesthetic polish<\/td>\n<td>Midjourney app only<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n<h2><span class=\"ez-toc-section\" id=\"Which_model_should_you_build_on\"><\/span>Which model should you build on?<span class=\"ez-toc-section-end\"><\/span><\/h2>\n<ul>\n<li><strong>You want to self-host or fine-tune:<\/strong> FLUX, or the Stable Diffusion 3.5 ecosystem if you need the deepest tooling.<\/li>\n<li><strong>You need precise prompt-following and editing in an app:<\/strong> GPT-4o image generation via the OpenAI API.<\/li>\n<li><strong>You&#8217;re on Google Cloud and want photorealism:<\/strong> Imagen.<\/li>\n<li><strong>You just want the best-looking output and don&#8217;t need to build on it:<\/strong> Midjourney, used through its app.<\/li>\n<li><strong>You need guaranteed clean licensing:<\/strong> Adobe Firefly&#8217;s model, which is trained on licensed data.<\/li>\n<\/ul>\n<p>For most developers in 2026, the decision is simple: use FLUX (or Stable Diffusion) when you need control, ownership, privacy, and no per-image cost; use a closed API model when you need top-tier instruction-following or photorealism and don&#8217;t mind paying per call.<\/p>\n<h2><span class=\"ez-toc-section\" id=\"Open_vs_closed_the_real_trade-off\"><\/span>Open vs closed: the real trade-off<span class=\"ez-toc-section-end\"><\/span><\/h2>\n<p>Open models (FLUX, Stable Diffusion) give you ownership: run them offline, fine-tune them on your own data, embed them in a product, pay nothing per image, and keep all data private. The cost is that you manage the infrastructure and the quality ceiling depends on your effort.<\/p>\n<p>Closed models (GPT-4o, Imagen, Midjourney&#8217;s) give you polish and convenience with zero infrastructure \u2014 but you rent access, pay per use, can&#8217;t customize the model itself, and send your prompts to a third party. Neither is universally better; the choice depends on whether control or convenience matters more for your use case.<\/p>\n<p><!--ai-enriched--><\/p>\n<h2><span class=\"ez-toc-section\" id=\"What_it_costs_to_generate_images_at_scale\"><\/span>What it costs to generate images at scale<span class=\"ez-toc-section-end\"><\/span><\/h2>\n<p>The model-quality debate matters less once you are generating thousands of images, where the pricing structure decides your bill more than the aesthetics do. The leading options split into three cost models, and the cheapest depends entirely on volume.<\/p>\n<p><strong>Per-image APIs<\/strong> are the default for products and automation. You pay only for what you generate, with no subscription: Flux 2 Pro runs roughly $0.05&ndash;$0.08 per image on hosted providers like fal.ai and Replicate, Stable Diffusion endpoints are cheaper still at a few cents, and OpenAI&#8217;s GPT Image and Google&#8217;s Imagen bill per image through their APIs. This scales linearly &mdash; ideal for spiky or low volume, expensive at high volume.<\/p>\n<p><strong>Subscriptions<\/strong> suit heavy, hands-on creative work. Midjourney has no official public API and charges roughly $10&ndash;$60\/month for effectively high-volume generation through its web app and Discord; for an artist iterating all day, a flat fee beats per-image metering. Ideogram and others offer similar free-plus-paid tiers.<\/p>\n<p><strong>Self-hosting<\/strong> is the zero-marginal-cost route for open-weight models. Stable Diffusion and the open Flux variants run on your own GPU, so after the hardware outlay each image is effectively just electricity &mdash; the economics that win at very high volume or where data must stay private. The trade-offs are setup effort, a capable GPU (a 12&ndash;24&nbsp;GB card for comfortable use), and a licensing caveat: some open checkpoints, such as the larger Flux <em>dev<\/em> weights, are non-commercial unless you buy a separate license.<\/p>\n<p>Regra pr\u00e1tica: <strong>per-image APIs for products and low volume, a subscription for daily creative iteration, and self-hosting once your volume or privacy needs make a GPU pay for itself.<\/strong><\/p>\n<h2><span class=\"ez-toc-section\" id=\"FAQ\"><\/span>Perguntas frequentes<span class=\"ez-toc-section-end\"><\/span><\/h2>\n<h3>What is the difference between diffusion and autoregressive image models?<\/h3>\n<p>Diffusion models generate an image by starting from noise and progressively refining it \u2014 they excel at texture and visual quality. Autoregressive models generate the image as a sequence of tokens, like a language model generates words \u2014 they excel at following precise instructions and rendering text. Many modern systems combine both approaches.<\/p>\n<h3>What is the best open-source image generation model?<\/h3>\n<p>FLUX is widely considered the best open-weight image model in 2026 \u2014 strong quality, good prompt adherence, and downloadable weights you can run and fine-tune. The Stable Diffusion 3.5 ecosystem remains the most mature for customization and community tooling.<\/p>\n<h3>Can I run image generation models on my own computer?<\/h3>\n<p>Yes \u2014 open models like FLUX and Stable Diffusion can run on a consumer GPU with enough VRAM (generally 8\u201312 GB or more, depending on the model variant). Closed models like GPT-4o image generation, Imagen, and Midjourney&#8217;s model cannot be self-hosted; they&#8217;re available only through their providers.<\/p>\n<h3>Which image model is best for a startup or product?<\/h3>\n<p>For control, privacy, and no per-image cost, build on FLUX or Stable Diffusion and host it yourself. For the best prompt precision with no infrastructure to manage, use the GPT-4o image API. Many products use both: an open model for bulk generation and a closed API for high-precision cases.<\/p>\n<h3>Why can&#8217;t diffusion models render text well?<\/h3>\n<p>Diffusion models shape the whole image at once rather than reasoning symbol by symbol, so exact letterforms often come out garbled. Newer models \u2014 and autoregressive architectures in particular \u2014 have improved text rendering significantly, and tools like Ideogram are specifically tuned to get text right.<\/p>\n<h3>How much does it cost to generate an AI image?<\/h3>\n<p>It depends on the route. Hosted per-image APIs are the clearest: Flux 2 Pro is around $0.05&ndash;$0.08 per image and Stable Diffusion endpoints are a few cents, while OpenAI&#8217;s GPT Image and Google&#8217;s Imagen bill per image at broadly comparable rates. Midjourney instead charges a roughly $10&ndash;$60 monthly subscription for high-volume use rather than per image. If you self-host an open model on your own GPU, the per-image cost is effectively just electricity.<\/p>\n<h3>Is it cheaper to self-host or use an API?<\/h3>\n<p>Self-hosting wins at high, steady volume; APIs win for low or spiky usage. A hosted API has zero upfront cost and you pay per image, which is ideal until your monthly bill exceeds what a capable GPU would cost. Running an open model like Stable Diffusion or Flux locally front-loads the hardware spend but drops the marginal cost per image to near zero, and keeps your prompts and outputs private. Estimate your monthly image volume and compare it against both before committing.<\/p>\n<h3>Can I use AI-generated images commercially?<\/h3>\n<p>Usually yes on paid tiers, but read the specific license. Midjourney grants commercial rights on any paid plan; OpenAI and Google permit commercial use of API output; Flux is cleared for commercial use through its API and the Apache-licensed <em>klein<\/em> weights, but the larger open <em>dev<\/em> checkpoint is non-commercial unless you buy a self-hosted license. A separate caveat applies everywhere: under current US guidance a purely AI-generated image generally cannot be copyrighted, so you are licensed to use it but may be unable to stop others from copying an unmodified output.<\/p>\n<h2><span class=\"ez-toc-section\" id=\"Bottom_line\"><\/span>Conclus\u00e3o<span class=\"ez-toc-section-end\"><\/span><\/h2>\n<p>Behind every image app is a model, and in 2026 the model landscape splits cleanly. <strong>FLUX<\/strong> e o <strong>Stable Diffusion<\/strong> ecosystem own the open side \u2014 choose them for control, customization, privacy, and zero per-image cost. <strong>GPT-4o image generation<\/strong>, <strong>Imagen<\/strong>, e <strong>Midjourney&#8217;s model<\/strong> own the closed side \u2014 choose them for polish, precision, and convenience without infrastructure.<\/p>\n<p>If you&#8217;re building, start with FLUX and add a closed API only where you need its specific strengths. If you&#8217;re just generating images, you&#8217;re really choosing an app \u2014 and our <a href=\"\/pt\/top-ai-image-generators-2026\/\">best AI image generators guide<\/a> covers that decision in full.<\/p>\n<p><!--related-block--><\/p>\n<div class=\"convly-related\">\n<h2><span class=\"ez-toc-section\" id=\"Related_articles\"><\/span>Artigos relacionados<span class=\"ez-toc-section-end\"><\/span><\/h2>\n<ul>\n<li><a href=\"https:\/\/convly.ai\/pt\/veo-3-vs-kling-3-for-ai-video-2026\/\">Veo 3.1 vs Kling 3.0 para v\u00eddeos com IA em 2026: qual deles vence em realismo?<\/a><\/li>\n<li><a href=\"https:\/\/convly.ai\/pt\/best-cloud-gpu-providers-for-ai-2026\/\">Melhores provedores de GPU em nuvem para IA em 2026: RunPod, Lambda, Vast, Together e Replicate<\/a><\/li>\n<li><a href=\"https:\/\/convly.ai\/pt\/ai-translation-tools-compared\/\">The Best AI Translation Tools in 2026: DeepL vs Google vs ChatGPT<\/a><\/li>\n<li><a href=\"https:\/\/convly.ai\/pt\/ai-music-generators-suno-vs-udio\/\">AI Music Generators in 2026: Suno vs Udio (Hands-On Review)<\/a><\/li>\n<\/ul>\n<\/div>","protected":false},"excerpt":{"rendered":"<p>Behind every image app is a model. This guide explains how 2026&#8217;s image generation models actually work \u2014 diffusion vs autoregressive \u2014 and compares the major model families for builders and power users.<\/p>","protected":false},"author":0,"featured_media":72,"comment_status":"closed","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"site-sidebar-layout":"default","site-content-layout":"","ast-site-content-layout":"default","site-content-style":"default","site-sidebar-style":"default","ast-global-header-display":"","ast-banner-title-visibility":"","ast-main-header-display":"","ast-hfb-above-header-display":"","ast-hfb-below-header-display":"","ast-hfb-mobile-header-display":"","site-post-title":"","ast-breadcrumbs-content":"","ast-featured-img":"","footer-sml-layout":"","ast-disable-related-posts":"","theme-transparent-header-meta":"","adv-header-id-meta":"","stick-header-meta":"","header-above-stick-meta":"","header-main-stick-meta":"","header-below-stick-meta":"","astra-migrate-meta-layouts":"default","ast-page-background-enabled":"default","ast-page-background-meta":{"desktop":{"background-color":"var(--ast-global-color-5)","background-image":"","background-repeat":"repeat","background-position":"center center","background-size":"auto","background-attachment":"scroll","background-type":"","background-media":"","overlay-type":"","overlay-color":"","overlay-opacity":"","overlay-gradient":""},"tablet":{"background-color":"","background-image":"","background-repeat":"repeat","background-position":"center center","background-size":"auto","background-attachment":"scroll","background-type":"","background-media":"","overlay-type":"","overlay-color":"","overlay-opacity":"","overlay-gradient":""},"mobile":{"background-color":"","background-image":"","background-repeat":"repeat","background-position":"center center","background-size":"auto","background-attachment":"scroll","background-type":"","background-media":"","overlay-type":"","overlay-color":"","overlay-opacity":"","overlay-gradient":""}},"ast-content-background-meta":{"desktop":{"background-color":"var(--ast-global-color-4)","background-image":"","background-repeat":"repeat","background-position":"center center","background-size":"auto","background-attachment":"scroll","background-type":"","background-media":"","overlay-type":"","overlay-color":"","overlay-opacity":"","overlay-gradient":""},"tablet":{"background-color":"var(--ast-global-color-4)","background-image":"","background-repeat":"repeat","background-position":"center center","background-size":"auto","background-attachment":"scroll","background-type":"","background-media":"","overlay-type":"","overlay-color":"","overlay-opacity":"","overlay-gradient":""},"mobile":{"background-color":"var(--ast-global-color-4)","background-image":"","background-repeat":"repeat","background-position":"center center","background-size":"auto","background-attachment":"scroll","background-type":"","background-media":"","overlay-type":"","overlay-color":"","overlay-opacity":"","overlay-gradient":""}},"footnotes":""},"categories":[5],"tags":[395,392,393,391,394],"class_list":["post-71","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-ai-tools","tag-ai-model-comparison","tag-diffusion-models","tag-flux-model","tag-image-generation-models","tag-stable-diffusion-3-5"],"_links":{"self":[{"href":"https:\/\/convly.ai\/pt\/wp-json\/wp\/v2\/posts\/71","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/convly.ai\/pt\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/convly.ai\/pt\/wp-json\/wp\/v2\/types\/post"}],"replies":[{"embeddable":true,"href":"https:\/\/convly.ai\/pt\/wp-json\/wp\/v2\/comments?post=71"}],"version-history":[{"count":3,"href":"https:\/\/convly.ai\/pt\/wp-json\/wp\/v2\/posts\/71\/revisions"}],"predecessor-version":[{"id":1036,"href":"https:\/\/convly.ai\/pt\/wp-json\/wp\/v2\/posts\/71\/revisions\/1036"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/convly.ai\/pt\/wp-json\/wp\/v2\/media\/72"}],"wp:attachment":[{"href":"https:\/\/convly.ai\/pt\/wp-json\/wp\/v2\/media?parent=71"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/convly.ai\/pt\/wp-json\/wp\/v2\/categories?post=71"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/convly.ai\/pt\/wp-json\/wp\/v2\/tags?post=71"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}