{"id":5857,"date":"2026-07-28T09:00:15","date_gmt":"2026-07-28T16:00:15","guid":{"rendered":"https:\/\/www.visla.us\/blog\/?p=5857"},"modified":"2026-08-14T11:01:22","modified_gmt":"2026-08-14T18:01:22","slug":"how-to-prompt-sora-2","status":"publish","type":"post","link":"https:\/\/www.visla.us\/blog\/guides\/how-to-prompt-sora-2\/","title":{"rendered":"How to Prompt Sora 2: What Still Works After Its Discontinuation"},"content":{"rendered":"\n<p class=\"wp-block-paragraph\"><em>Updated 7\/28\/2026<\/em><\/p>\n\n\n\n<div class=\"wp-block-group has-base-2-background-color has-background has-global-padding is-layout-constrained wp-container-core-group-is-layout-34a480d6 wp-block-group-is-layout-constrained\" style=\"border-radius:10px;padding-top:var(--wp--preset--spacing--20);padding-right:var(--wp--preset--spacing--20);padding-bottom:var(--wp--preset--spacing--20);padding-left:var(--wp--preset--spacing--20)\">\n<p class=\"is-style-asterisk wp-block-paragraph\"><strong>Quick answer: Can you still use and prompt Sora 2?<\/strong><\/p>\n\n\n\n<p class=\"wp-block-paragraph\">Sora 2\u2019s consumer app and website shut down on April 26, 2026, and OpenAI will discontinue the Sora API on September 24, 2026. The API remains technically available during that migration period, but both Sora 2 models are deprecated, and Visla has already moved away from them. The good news is that Sora 2\u2019s core prompting principles still apply to current AI video models. This guide explains what happened, what Sora\u2019s shutdown may tell us about the market, and how to prompt today\u2019s models and Visla\u2019s AI Director Mode.<\/p>\n<\/div>\n\n\n\n<figure class=\"wp-block-embed is-type-video is-provider-youtube wp-block-embed-youtube wp-embed-aspect-16-9 wp-has-aspect-ratio\"><div class=\"wp-block-embed__wrapper\">\n<iframe loading=\"lazy\" title=\"Why OpenAI Shut Down Sora 2\u2014and What Comes Next for AI Video\" width=\"500\" height=\"281\" src=\"https:\/\/www.youtube.com\/embed\/Aoh4ZlFw4iI?feature=oembed\" frameborder=\"0\" allow=\"accelerometer; autoplay; clipboard-write; encrypted-media; gyroscope; picture-in-picture; web-share\" referrerpolicy=\"strict-origin-when-cross-origin\" allowfullscreen><\/iframe>\n<\/div><figcaption class=\"wp-element-caption\"><em>Video made using Visla.<\/em><\/figcaption><\/figure>\n\n\n\n<h2 class=\"wp-block-heading\">What happened to Sora 2?<\/h2>\n\n\n\n<p class=\"wp-block-paragraph\">When <a href=\"https:\/\/openai.com\/index\/sora-2\/\" target=\"_blank\" rel=\"noreferrer noopener\">OpenAI launched Sora 2<\/a> in September 2025, it introduced a video-generation model and a social app built around it. Users could create clips, remix other people\u2019s work, add themselves or friends as recurring characters, and browse a personalized feed. OpenAI presented Sora as a social way to experience AI video rather than a conventional production platform.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">That experiment didn\u2019t last long. According to <a href=\"https:\/\/help.openai.com\/en\/articles\/20001152-what-to-know-about-the-sora-discontinuation\" target=\"_blank\" rel=\"noreferrer noopener\">OpenAI\u2019s official Sora discontinuation notice<\/a>, the web and mobile products closed on April 26, 2026. The API remains available until September 24, but it\u2019s deprecated and has no announced replacement. Developers can maintain existing integrations during the migration period, but they shouldn\u2019t build new long-term workflows around Sora 2.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">Visla has already removed Sora 2 from its active model lineup. The old Sora-specific instructions in this article are therefore historical, even though the broader prompting advice remains useful.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">OpenAI hasn\u2019t published a detailed financial postmortem. Reporting on <a href=\"https:\/\/www.theverge.com\/ai-artificial-intelligence\/902368\/openai-sora-dead-ai-video-generation-competition\" target=\"_blank\" rel=\"noreferrer noopener\">why OpenAI discontinued Sora<\/a> points to heavy compute use, declining app momentum, and a broader shift toward coding, enterprise tools, agents, and world-simulation research.<\/p>\n\n\n\n<h2 class=\"wp-block-heading\">The expensive economics behind disposable AI video<\/h2>\n\n\n\n<p class=\"wp-block-paragraph\">Sora\u2019s economic mismatch was straightforward. OpenAI was trying to run a social feed where producing much of the content consumed scarce and expensive compute.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">TikTok, Instagram, and YouTube have to attract creators, recommend posts, moderate content, and keep users returning. Their users, however, usually absorb the work and cost of filming and editing the videos. Sora still had to solve the same platform problems while paying the computational cost whenever someone generated a clip.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">The social part couldn\u2019t be treated as an afterthought. A <a href=\"https:\/\/arxiv.org\/abs\/2507.04545\" target=\"_blank\" rel=\"noreferrer noopener\">2025 study of network effects on Facebook, Instagram, LinkedIn, and X<\/a> found that connections to other users explained between 20% and 34% of the value people received from those platforms. Sora had remixing and reusable characters, but it also needed creators, viewers, and meaningful social connections to turn an impressive generation tool into a network people reopened every day.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">OpenAI has proved that it can build widely used utilities and developer products. A creator-first social network asks for different strengths, including creator relations, feed operations, cultural judgment, and moderation at social-media scale. Sora had to develop those capabilities while competing against established platforms and improving an expensive video model.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">The initial download surge also faded quickly. Sensor Tower data <a href=\"https:\/\/www.theverge.com\/ai-artificial-intelligence\/902368\/openai-sora-dead-ai-video-generation-competition\" target=\"_blank\" rel=\"noreferrer noopener\">reported by The Verge<\/a> showed 6.1 million worldwide downloads in November 2025, followed by 3.2 million in December, 2.1 million in January, 1.4 million in February, and about 1.1 million in March before the shutdown announcement. Downloads aren\u2019t the same as active-user retention, but the pattern suggests that launch attention wasn\u2019t turning into sustained growth.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">OpenAI could build a competitive video model. The harder task was turning expensive generations into a social habit strong enough to justify the compute. Video quality alone couldn\u2019t solve that problem.<\/p>\n\n\n\n<h2 class=\"wp-block-heading\">What Sora 2\u2019s shutdown means for AI video<\/h2>\n\n\n\n<p class=\"wp-block-paragraph\">Sora\u2019s failure as a social product doesn\u2019t mean AI video lacks value. It suggests that the strongest use cases may begin with a practical job instead of the desire to generate a surprising clip.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">Coding provides the clearest comparison. A useful coding tool doesn\u2019t stop after producing an impressive function. The user tests the code, revises it, connects it to an existing project, and deploys it. The value comes from working software.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">Businesses need the same thing from AI video. A training team needs an accurate video it can update when a policy changes. A marketing team needs approved language, correct branding, and versions for different channels. A sales team may need an explainer that combines generated footage with product screenshots, subtitles, and a call to action.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">Those needs extend beyond clip generation. Teams have to preserve source material, maintain visual continuity, review scenes, correct mistakes, manage assets, and deliver the finished video in the right format.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">That helps explain why companies are spending heavily on specialized AI applications. <a href=\"https:\/\/menlovc.com\/perspective\/2025-the-state-of-generative-ai-in-the-enterprise\/\" target=\"_blank\" rel=\"noreferrer noopener\">Menlo Ventures\u2019 2025 enterprise AI report<\/a> estimated that the application layer captured $19 billion, more than half of enterprise generative AI spending. Coding was the largest departmental category, but marketing, customer support, IT, design, and HR also attracted meaningful investment.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">The model still matters. Better models improve motion, prompt adherence, audio, and visual consistency. But a business application becomes useful when it organizes those capabilities around a repeatable task.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">For AI video, that means turning source material into a planned project, generating the footage that\u2019s actually needed, and leaving the team with something it can edit, approve, and publish.<\/p>\n\n\n\n<h2 class=\"wp-block-heading\">How to write a strong AI video prompt<\/h2>\n\n\n\n<p class=\"wp-block-paragraph\">A good AI video prompt describes what a camera should capture. It gives the model visible actions and specific production direction instead of relying on broad requests such as \u201cmake it cinematic.\u201d<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">Most strong prompts cover five areas:<\/p>\n\n\n\n<ol start=\"1\" class=\"wp-block-list\">\n<li><strong>Subject and setting:<\/strong> Identify who or what appears, where the scene takes place, the time of day, and any important objects.<\/li>\n\n\n\n<li><strong>Action:<\/strong> Describe movement in the order it should happen. Keep the sequence short enough to fit the clip.<\/li>\n\n\n\n<li><strong>Camera:<\/strong> Specify framing, angle, lens behavior, and movement.<\/li>\n\n\n\n<li><strong>Lighting and style:<\/strong> Describe the direction and quality of the light, color palette, texture, and visual medium.<\/li>\n\n\n\n<li><strong>Audio:<\/strong> When supported, separate dialogue, ambient noise, music, and sound effects.<\/li>\n<\/ol>\n\n\n\n<p class=\"wp-block-paragraph\">Google\u2019s <a href=\"https:\/\/cloud.google.com\/blog\/products\/ai-machine-learning\/ultimate-prompting-guide-for-veo-3-1\">Veo 3.1 prompting guide<\/a> uses a similar framework: cinematography, subject, action, context, and style or ambiance. It also recommends describing dialogue and sound directly when the model supports native audio.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">Here\u2019s a model-neutral example:<\/p>\n\n\n\n<blockquote class=\"wp-block-quote is-layout-flow wp-block-quote-is-layout-flow\">\n<p class=\"has-medium-font-size wp-block-paragraph\">A tired office worker sits alone at a kitchen table before sunrise, reviewing a printed training manual. Medium close-up at eye level. The camera slowly pushes forward as she highlights one sentence and smiles with relief. Cool blue window light fills the room, while a warm lamp illuminates the paper. Quiet room tone, distant traffic, and the soft scrape of a highlighter.<\/p>\n<\/blockquote>\n\n\n\n<figure class=\"wp-block-video\"><video height=\"1080\" style=\"aspect-ratio: 1920 \/ 1080;\" width=\"1920\" controls src=\"https:\/\/www.visla.us\/wp-content\/uploads\/2025\/10\/Female-Office-Worker-at-Kitchen-Table-Before-Sunrise.mp4\"><\/video><figcaption class=\"wp-element-caption\"><em>Video clip generated using Veo 3.1<\/em><\/figcaption><\/figure>\n\n\n\n<p class=\"wp-block-paragraph\">Treat the prompt as a brief for one shot or short sequence. Don\u2019t ask a single generation to cover several locations, costume changes, and conflicting camera movements. Break the idea into separate clips and assemble them during editing.<\/p>\n\n\n\n<h2 class=\"wp-block-heading\">Which prompting advice depends on the AI video model?<\/h2>\n\n\n\n<p class=\"wp-block-paragraph\">The basic visual language transfers across models, but different tools give users different kinds of control.<\/p>\n\n\n\n<h3 class=\"wp-block-heading\">Use conversational editing with Gemini Omni Flash<\/h3>\n\n\n\n<p class=\"wp-block-paragraph\"><a href=\"https:\/\/cloud.google.com\/blog\/products\/ai-machine-learning\/nano-banana-2-lite-and-gemini-omni-flash-available\" target=\"_blank\" rel=\"noreferrer noopener\">Gemini Omni Flash<\/a> supports video generation and conversational editing with text, image, and video inputs, plus native audio. You can generate a scene, inspect it, and then request a precise change without rebuilding the whole prompt.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">For example:<\/p>\n\n\n\n<blockquote class=\"wp-block-quote is-layout-flow wp-block-quote-is-layout-flow\">\n<p class=\"has-medium-font-size wp-block-paragraph\">Replace the presenter with the woman in the reference image. Keep the camera movement, office background, timing, dialogue, and ambient audio unchanged.<\/p>\n<\/blockquote>\n\n\n\n<p class=\"wp-block-paragraph\">This works best when the instruction distinguishes what should change from what must remain stable.<\/p>\n\n\n\n<h3 class=\"wp-block-heading\">Assign every Seedance 2.0 reference a role<\/h3>\n\n\n\n<p class=\"wp-block-paragraph\"><a href=\"https:\/\/seed.bytedance.com\/en\/blog\/official-launch-of-seedance-2-0\" target=\"_blank\" rel=\"noreferrer noopener\">Seedance 2.0<\/a> accepts text, images, video, and audio together. It can take a character\u2019s appearance from one reference, motion from another, and timing or atmosphere from an audio clip.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">Several unexplained references can create ambiguity. Tell the model exactly what to borrow:<\/p>\n\n\n\n<blockquote class=\"wp-block-quote is-layout-flow wp-block-quote-is-layout-flow\">\n<p class=\"has-medium-font-size wp-block-paragraph\">Use image one for the character\u2019s appearance and clothing. Use video one for her walking speed and gestures. Use image two for the office layout. Match the rhythm and ambient sound of the audio reference, but don\u2019t reproduce its dialogue.<\/p>\n<\/blockquote>\n\n\n\n<h3 class=\"wp-block-heading\">Direct Veo 3.1\u2019s sound and transitions<\/h3>\n\n\n\n<p class=\"wp-block-paragraph\"><a href=\"https:\/\/cloud.google.com\/blog\/products\/ai-machine-learning\/ultimate-prompting-guide-for-veo-3-1\" target=\"_blank\" rel=\"noreferrer noopener\">Veo 3.1<\/a> provides detailed controls for native audio, reference ingredients, and transitions between first and last frames. Separate visual action from audio direction:<\/p>\n\n\n\n<blockquote class=\"wp-block-quote is-layout-flow wp-block-quote-is-layout-flow\">\n<p class=\"has-medium-font-size wp-block-paragraph\">Medium two-shot of two employees standing beside a stalled conveyor belt. One looks at the control panel and says, \u201cThe emergency stop is still engaged.\u201d Ambient sound: low factory hum and distant machinery. SFX: a warning light clicks off.<\/p>\n<\/blockquote>\n\n\n\n<p class=\"wp-block-paragraph\">When using first and last frames, focus the prompt on how the subject and camera move between those fixed images.<\/p>\n\n\n\n<h3 class=\"wp-block-heading\">Focus on motion with Runway Gen-4.5<\/h3>\n\n\n\n<p class=\"wp-block-paragraph\"><a href=\"https:\/\/help.runwayml.com\/hc\/en-us\/articles\/46974685288467-Creating-with-Gen-4-5\" target=\"_blank\" rel=\"noreferrer noopener\">Runway Gen-4.5<\/a> supports text-to-video and image-to-video generation. For image-to-video, the starting image already establishes the subject, composition, lighting, and style. Use the prompt to describe what changes:<\/p>\n\n\n\n<blockquote class=\"wp-block-quote is-layout-flow wp-block-quote-is-layout-flow\">\n<p class=\"has-medium-font-size wp-block-paragraph\">The camera slowly tracks left. The presenter turns toward the display, raises her right hand, and points to the top chart. The display\u2019s bars rise one at a time. Her clothing and facial features remain unchanged.<\/p>\n<\/blockquote>\n\n\n\n<p class=\"wp-block-paragraph\">That gives the model useful motion and timing instructions instead of redescribing information it can already see.<\/p>\n\n\n\n<h2 class=\"wp-block-heading\">How to prompt AI Director Mode<\/h2>\n\n\n\n<p class=\"wp-block-paragraph\"><a href=\"https:\/\/www.visla.us\/ai-director-mode\" target=\"_blank\" rel=\"noreferrer noopener\">Visla\u2019s AI Director Mode<\/a> distributes prompting across the full video-production process. You give direction at the project level, refine recurring visual elements, and review each scene prompt before generating motion.<\/p>\n\n\n\n<h3 class=\"wp-block-heading\">1. Input: Brief the complete video<\/h3>\n\n\n\n<p class=\"wp-block-paragraph\">Start with an idea, script, webpage, PDF, or uploaded assets. Explain the intended audience, purpose, tone, length, and format. Identify any facts, quotes, or approved wording that must stay intact.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">At this stage, you\u2019re prompting the whole assignment:<\/p>\n\n\n\n<blockquote class=\"wp-block-quote is-layout-flow wp-block-quote-is-layout-flow\">\n<p class=\"has-medium-font-size wp-block-paragraph\">Create a 90-second training video for new warehouse employees. Explain the three steps required before operating a pallet jack. Use direct, reassuring language and preserve the safety instructions from the uploaded PDF.<\/p>\n<\/blockquote>\n\n\n\n<h3 class=\"wp-block-heading\">2. Planning: Define recurring elements<\/h3>\n\n\n\n<p class=\"wp-block-paragraph\">AI Director Mode reviews the input and suggests characters, objects, environments, and visual style. You can add, remove, or edit those elements before Visla creates the storyboard.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">This is where you decide what must remain consistent across the video, including the presenter, workplace, equipment, product appearance, and overall visual approach.<\/p>\n\n\n\n<h3 class=\"wp-block-heading\">3. Generation: Prompt individual scenes<\/h3>\n\n\n\n<p class=\"wp-block-paragraph\">Visla shows an AI-generated image and an editable prompt for each scene. Review the subject, visible action, framing, camera movement, lighting, and sound. Refine the image and prompt before generating a full AI video clip.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">You don\u2019t need motion in every scene. A diagram, screenshot, stock clip, or storyboard image may communicate some information more clearly. Generate AI video where movement contributes to the message.<\/p>\n\n\n\n<h3 class=\"wp-block-heading\">4. Output: Edit the finished project<\/h3>\n\n\n\n<p class=\"wp-block-paragraph\">Continue in <a href=\"https:\/\/www.visla.us\/scene-based-video-editing\" target=\"_blank\" rel=\"noreferrer noopener\">Visla\u2019s Scene-Based Editor<\/a>. Adjust pacing, replace weak visuals, update the script or voiceover, add branding, and review the full video for continuity.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">At this point, the main job isn\u2019t prompting. It\u2019s finishing the video your audience will receive.<\/p>\n\n\n\n<div class=\"wp-block-buttons is-layout-flex wp-block-buttons-is-layout-flex\">\n<div class=\"wp-block-button\"><a class=\"wp-block-button__link wp-element-button\" href=\"https:\/\/app.visla.us\/signup\" target=\"_blank\" rel=\"noreferrer noopener\">Try AI Director Mode for yourself<\/a><\/div>\n<\/div>\n\n\n\n<h2 class=\"wp-block-heading\">FAQ<\/h2>\n\n\n\n<div class=\"schema-faq wp-block-yoast-faq-block\"><div class=\"schema-faq-section\" id=\"faq-question-1761003139370\"><strong class=\"schema-faq-question\">Can a business copyright an AI-generated video?<\/strong> <p class=\"schema-faq-answer\">It depends on how much human authorship went into the finished video. The U.S. Copyright Office says purely AI-generated material generally can\u2019t receive copyright protection, but human-created elements can. That may include the script, narration, editing, graphics, arrangement of scenes, and meaningful changes made to generated footage. Simply writing a prompt may not provide enough creative control on its own. Businesses should also confirm that they have permission to use any reference images, music, voices, logos, or personal likenesses included in the video. Copyright rules differ by country, so companies producing high-value commercial work should get advice for the markets where they plan to publish it.<\/p> <\/div> <div class=\"schema-faq-section\" id=\"faq-question-1761064291846\"><strong class=\"schema-faq-question\">Do you have to label videos made with AI?<\/strong> <p class=\"schema-faq-answer\">Major social platforms generally require disclosure when AI creates or meaningfully alters realistic footage. YouTube requires creators to disclose realistic scenes that didn\u2019t happen, altered footage of real events, or depictions of real people saying and doing things they never did. TikTok similarly requires labels for realistic AI-generated images, audio, and video. Meta requires its disclosure tool for organic posts containing photorealistic AI video or realistic-sounding synthetic audio. Minor production assistance, such as generating captions, repairing audio, or adjusting lighting, usually doesn\u2019t require the same label. Check each platform\u2019s rules before publishing because the requirements and available labels continue to change.<\/p> <\/div> <div class=\"schema-faq-section\" id=\"faq-question-1761064303174\"><strong class=\"schema-faq-question\">Can you monetize AI-generated videos on YouTube?<\/strong> <p class=\"schema-faq-answer\">Yes. YouTube doesn\u2019t automatically disqualify a video because it contains AI-generated footage, voices, or other elements. It also says that using its AI disclosure setting won\u2019t limit a video\u2019s audience or affect its eligibility to earn money. The larger risk is publishing repetitive or mass-produced material with little original value. YouTube\u2019s monetization rules classify that as \u201cinauthentic content,\u201d whether AI was involved or not. A channel is in a stronger position when each video includes original research, commentary, storytelling, editing, instruction, or another clear human contribution instead of repeating the same template at scale.<\/p> <\/div> <\/div>\n","protected":false},"excerpt":{"rendered":"<p>Updated 7\/28\/2026 Quick answer: Can you still use and prompt Sora 2? Sora 2\u2019s consumer app and website shut down on April 26, 2026, and OpenAI will discontinue the Sora API on September 24, 2026. The API remains technically available during that migration period, but both Sora 2 models are deprecated, and Visla has already [&hellip;]<\/p>\n","protected":false},"author":9,"featured_media":7482,"comment_status":"closed","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"_monsterinsights_skip_tracking":false,"slim_seo":{"description":"Learn which Sora 2 prompting techniques still work, what changed after its 2026 discontinuation, and how to adapt prompts for current AI video tools.","title":"How to Prompt Sora 2: What Still Works After Its Discontinuation - The Visla Blog"},"footnotes":""},"categories":[25],"tags":[31],"class_list":["post-5857","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-guides","tag-ai-video"],"_links":{"self":[{"href":"https:\/\/www.visla.us\/blog\/wp-json\/wp\/v2\/posts\/5857","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.visla.us\/blog\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/www.visla.us\/blog\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/www.visla.us\/blog\/wp-json\/wp\/v2\/users\/9"}],"replies":[{"embeddable":true,"href":"https:\/\/www.visla.us\/blog\/wp-json\/wp\/v2\/comments?post=5857"}],"version-history":[{"count":28,"href":"https:\/\/www.visla.us\/blog\/wp-json\/wp\/v2\/posts\/5857\/revisions"}],"predecessor-version":[{"id":7499,"href":"https:\/\/www.visla.us\/blog\/wp-json\/wp\/v2\/posts\/5857\/revisions\/7499"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/www.visla.us\/blog\/wp-json\/wp\/v2\/media\/7482"}],"wp:attachment":[{"href":"https:\/\/www.visla.us\/blog\/wp-json\/wp\/v2\/media?parent=5857"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.visla.us\/blog\/wp-json\/wp\/v2\/categories?post=5857"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.visla.us\/blog\/wp-json\/wp\/v2\/tags?post=5857"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}