{"id":10471,"date":"2025-07-31T11:12:50","date_gmt":"2025-07-31T11:12:50","guid":{"rendered":"https:\/\/www.d-id.com\/?p=10471"},"modified":"2026-02-18T12:59:27","modified_gmt":"2026-02-18T12:59:27","slug":"how-to-build-a-d%e2%80%91id-visual-agent-a-prompt%e2%80%91by%e2%80%91prompt-guide","status":"publish","type":"post","link":"https:\/\/www.d-id.com\/blog\/how-to-build-a-did-visual-agent-a-promptbyprompt-guide\/","title":{"rendered":"How to Build a D\u2011ID Visual Agent: A Prompt\u2011by\u2011Prompt Guide"},"content":{"rendered":"\n<h2 class=\"wp-block-heading\" id=\"h-what-are-visual-agents\"><strong>What Are Visual Agents?<\/strong><\/h2>\n\n\n\n<p class=\"has-text-align-left\">If you\u2019ve ever wished your chatbot could look you in the eye, smile, and hold a natural conversation, you\u2019re in the right place. D\u2011ID\u2019s Visual Agents make that possible. No cameras, no crews, just a few clicks (and the right prompts) in the Studio. <br><br><a href=\"https:\/\/www.d-id.com\/ai-agents\/\">Visual agents <\/a>are interactive AI avatars that are <em>live<\/em>, conversational, and powered by real\u2011time AI. They combine human presence (<a href=\"https:\/\/www.d-id.com\/blog\/ai-agents-vs-ai-avatars\/\">through avatars<\/a>) with AI intelligence (through live conversational models). <br><br>Click this <a href=\"https:\/\/studio.d-id.com\/agents\/share?id=v2_agt_EZWsVkEt&amp;utm_source=copy&amp;key=WjI5dloyeGxMVzloZFhSb01ud3hNREl6TXpFek5ETTVNemt5TnpZM01EUTJNVE02TjJ4QkxWSlJjMmhwZFVaeVpYUlFibmhuTjBkcw==\">link<\/a> to speak with Amber, a D-ID visual agent.<br><br>This guide walks you through creating your first visual agent, prompt by prompt. Whether you\u2019re welcoming website visitors, answering FAQs, or just showing off what\u2019s possible, you\u2019ll learn what each Studio field does, how to fill it, and how to get a visual agent that feels alive, on brand, and ready to engage.<br><\/p>\n\n\n\n<hr class=\"wp-block-separator has-alpha-channel-opacity\"\/>\n\n\n\n<h2 class=\"wp-block-heading\" id=\"h-tab-1-appearance-choosing-the-face-of-your-visual-agent\"><strong>Tab 1: Appearance &#8211; Choosing the Face of Your Visual Agent<\/strong><\/h2>\n\n\n\n<figure class=\"wp-block-image size-large is-resized\"><img decoding=\"async\" width=\"1024\" height=\"584\" src=\"https:\/\/www.d-id.com\/wp-content\/uploads\/2025\/07\/Selecting-your-visual-agents-appearance-1024x584.png\" alt=\"A user interface shows an avatar selection menu on the left with various avatars labeled \u201cPremium\u201d and a preview of a female avatar on the right in a chat setup.\" class=\"wp-image-10476\" style=\"width:1177px;height:auto\" srcset=\"https:\/\/www.d-id.com\/wp-content\/uploads\/2025\/07\/Selecting-your-visual-agents-appearance-1024x584.png 1024w, https:\/\/www.d-id.com\/wp-content\/uploads\/2025\/07\/Selecting-your-visual-agents-appearance-300x171.png 300w, https:\/\/www.d-id.com\/wp-content\/uploads\/2025\/07\/Selecting-your-visual-agents-appearance-768x438.png 768w, https:\/\/www.d-id.com\/wp-content\/uploads\/2025\/07\/Selecting-your-visual-agents-appearance-1536x875.png 1536w, https:\/\/www.d-id.com\/wp-content\/uploads\/2025\/07\/Selecting-your-visual-agents-appearance-2048x1167.png 2048w\" sizes=\"(max-width: 1024px) 100vw, 1024px\" \/><\/figure>\n\n\n\n<p>Your visual agent\u2019s appearance is its first impression. It\u2019s what makes users stop, pay attention, and feel like they\u2019re talking to a person, not just a piece of software.<\/p>\n\n\n\n<p>In D\u2011ID Studio, the Appearance field is where you select or create your agent&#8217;s avatar.<\/p>\n\n\n\n<h3 class=\"wp-block-heading\"><strong>Two Ways to Set the Appearance<\/strong><\/h3>\n\n\n\n<h5 class=\"wp-block-heading\"><strong>1. Stock Avatars<\/strong><\/h5>\n\n\n\n<ul class=\"wp-block-list\">\n<li><strong>What they are:<\/strong> A curated library of ready\u2011made digital people.<\/li>\n\n\n\n<li><strong>Best for:<\/strong> Quick setup, testing new agents, or use cases where the face doesn\u2019t need to match a specific brand personality.<\/li>\n\n\n\n<li><strong>Pros:<\/strong>\n<ul class=\"wp-block-list\">\n<li>Instant access &#8211; pick and go.<\/li>\n\n\n\n<li>Wide variety of demographics and styles.<\/li>\n\n\n\n<li>No production work required.<\/li>\n\n\n\n<li>Studio-quality trained on professional actors<\/li>\n<\/ul>\n<\/li>\n\n\n\n<li><strong>Cons:<\/strong>\n<ul class=\"wp-block-list\">\n<li>Not unique to your brand.<\/li>\n<\/ul>\n<\/li>\n<\/ul>\n\n\n\n<h5 class=\"wp-block-heading\"><strong>2. Custom Avatars<\/strong><\/h5>\n\n\n\n<ul class=\"wp-block-list\">\n<li><strong>What they are:<\/strong> Your own uploaded images or videos turned into an avatar.<\/li>\n\n\n\n<li><strong>Best for:<\/strong> Brand\u2011aligned Agents (e.g., spokesperson, team member, influencer).<\/li>\n\n\n\n<li><strong>Pros:<\/strong>\n<ul class=\"wp-block-list\">\n<li>Fully unique to you.<\/li>\n\n\n\n<li>Builds stronger brand familiarity.<\/li>\n<\/ul>\n<\/li>\n\n\n\n<li><strong>Cons:<\/strong>\n<ul class=\"wp-block-list\">\n<li>Requires you to create or source media.<\/li>\n\n\n\n<li>Premium+ tiers required for video-based uploads.<\/li>\n<\/ul>\n<\/li>\n<\/ul>\n\n\n\n<h3 class=\"wp-block-heading\"><strong>Two Formats for Avatars<\/strong><\/h3>\n\n\n\n<p>Regardless of whether you choose stock or custom, you can pick the format:<\/p>\n\n\n\n<figure class=\"wp-block-image size-medium\"><img decoding=\"async\" width=\"300\" height=\"187\" src=\"https:\/\/www.d-id.com\/wp-content\/uploads\/2025\/07\/Creating-a-new-avatar-for-your-visual-agent-300x187.png\" alt=\"Screenshot showing two options for creating an avatar: &quot;Create with a photo&quot; using a headshot, and &quot;Create with a video&quot; for higher quality, both featuring a woman in glasses and a striped shirt.\" class=\"wp-image-10477\" srcset=\"https:\/\/www.d-id.com\/wp-content\/uploads\/2025\/07\/Creating-a-new-avatar-for-your-visual-agent-300x187.png 300w, https:\/\/www.d-id.com\/wp-content\/uploads\/2025\/07\/Creating-a-new-avatar-for-your-visual-agent-1024x639.png 1024w, https:\/\/www.d-id.com\/wp-content\/uploads\/2025\/07\/Creating-a-new-avatar-for-your-visual-agent-768x479.png 768w, https:\/\/www.d-id.com\/wp-content\/uploads\/2025\/07\/Creating-a-new-avatar-for-your-visual-agent.png 1086w\" sizes=\"(max-width: 300px) 100vw, 300px\" \/><\/figure>\n\n\n\n<h5 class=\"wp-block-heading\"><strong>Photo\u2011Based (Standard)<\/strong><\/h5>\n\n\n\n<ul class=\"wp-block-list\">\n<li><strong>How it works:<\/strong> Uses a single still image to animate speech and expression.<\/li>\n\n\n\n<li><strong>Best for:<\/strong> Fast performance, lightweight interactions, simple informational Agents.<\/li>\n<\/ul>\n\n\n\n<h5 class=\"wp-block-heading\"><strong>Video\u2011Based (Premium \/ Premium+)<\/strong><\/h5>\n\n\n\n<ul class=\"wp-block-list\">\n<li><strong>How it works:<\/strong> Uses a short video clip for richer animation, more natural expressions, and subtle movements.<\/li>\n\n\n\n<li><strong>Best for:<\/strong> High\u2011impact experiences like sales demos, high\u2011touch customer service, or brand representation.<\/li>\n<\/ul>\n\n\n\n<p class=\"has-background\" style=\"background-color:#ff882e\"><strong>Pro Tip<\/strong>: <em>If your Agent is customer\u2011facing or plays a prominent role on your site\/app, <strong>invest in Premium+ custom avatars<\/strong>. <\/em><\/p>\n\n\n\n<hr class=\"wp-block-separator has-alpha-channel-opacity\"\/>\n\n\n\n<h2 class=\"wp-block-heading\" id=\"h-tab-2-agent-details-amp-preview-mode-define-how-your-visual-agent-acts\"><strong>Tab 2: Agent Details &amp; Preview Mode<\/strong> &#8211; Define How Your Visual Agent Acts<\/h2>\n\n\n\n<figure class=\"wp-block-image size-large\"><img decoding=\"async\" width=\"1024\" height=\"593\" src=\"https:\/\/www.d-id.com\/wp-content\/uploads\/2025\/07\/Creating-a-visual-agent-details-tab-1024x593.png\" alt=\"Screenshot of an AI agent setup page showing options to select name, language, voice, personality, and a preview of the virtual agent on the right side.\" class=\"wp-image-10478\" srcset=\"https:\/\/www.d-id.com\/wp-content\/uploads\/2025\/07\/Creating-a-visual-agent-details-tab-1024x593.png 1024w, https:\/\/www.d-id.com\/wp-content\/uploads\/2025\/07\/Creating-a-visual-agent-details-tab-300x174.png 300w, https:\/\/www.d-id.com\/wp-content\/uploads\/2025\/07\/Creating-a-visual-agent-details-tab-768x445.png 768w, https:\/\/www.d-id.com\/wp-content\/uploads\/2025\/07\/Creating-a-visual-agent-details-tab-1536x890.png 1536w, https:\/\/www.d-id.com\/wp-content\/uploads\/2025\/07\/Creating-a-visual-agent-details-tab-2048x1187.png 2048w\" sizes=\"(max-width: 1024px) 100vw, 1024px\" \/><\/figure>\n\n\n\n<p>Once your visual agent has a face, the next step is to give it a personality framework, the key details that shape how it\u2019s perceived. You\u2019ll also notice a window appear on the right side of the Studio. This is Preview Mode, your real\u2011time testing space. As you fill out the fields on this tab, the panel on the right lets you chat with your visual agent and see how your inputs affect its responses in real time. In preview, the visual agent won\u2019t be animated, but it will respond in text so you can test tone, style, and behavior before going live.<\/p>\n\n\n\n<h3 class=\"wp-block-heading\"><strong>Agent Name<\/strong><\/h3>\n\n\n\n<ul class=\"wp-block-list\">\n<li><strong>What it does:<\/strong> This is the name displayed to users during interaction.<\/li>\n\n\n\n<li><strong>Best practice:<\/strong>\n<ul class=\"wp-block-list\">\n<li>Keep it short, friendly, and easy to pronounce.<\/li>\n\n\n\n<li>Use first names only (\u201cAmber,\u201d \u201cAlex,\u201d \u201cEmma\u201d) for accessibility.<\/li>\n\n\n\n<li>Avoid quirky or joke names unless they fit your brand tone.<\/li>\n<\/ul>\n<\/li>\n\n\n\n<li><strong>Why it matters:<\/strong> The name is the first anchor point for building rapport, it\u2019s small but powerful.<\/li>\n<\/ul>\n\n\n\n<h3 class=\"wp-block-heading\"><strong>Language &amp; Voice<\/strong><\/h3>\n\n\n\n<ul class=\"wp-block-list\">\n<li><strong>What it does:<\/strong> Sets how your Agent sounds and in which language(s) it communicates.<\/li>\n\n\n\n<li><strong>Best practice:<\/strong>\n<ul class=\"wp-block-list\">\n<li>Match your audience\u2019s primary language.<\/li>\n\n\n\n<li>Pick a voice that fits the Persona, warm and approachable for casual interactions, calm and professional for support roles.<\/li>\n\n\n\n<li>Stick with one voice per Agent for consistency.<\/li>\n<\/ul>\n<\/li>\n\n\n\n<li><strong>Why it matters:<\/strong> Voice and language shape tone, clarity, and trust in every conversation.<\/li>\n<\/ul>\n\n\n\n<h3 class=\"wp-block-heading\" id=\"h-role\"><strong>Role<\/strong><\/h3>\n\n\n\n<ul class=\"wp-block-list\">\n<li><strong>What it does:<\/strong> Defines the visual agent\u2019s \u201cjob description\u201d in a single sentence.<\/li>\n\n\n\n<li><strong>Best practice:<\/strong>\n<ul class=\"wp-block-list\">\n<li>Format as \u201cYou are [name], a [tone\/role] who [main function].\u201d<\/li>\n\n\n\n<li>Be specific. Avoid vague roles like \u201cAI assistant.\u201d<\/li>\n\n\n\n<li>Example: \u201cYou are Chloe, a friendly customer support specialist who helps users troubleshoot and set up our product.\u201d<\/li>\n<\/ul>\n<\/li>\n\n\n\n<li><strong>Why it matters:<\/strong> Role sets the scope of interaction, keeping the visual agent focused and on-brand.<\/li>\n<\/ul>\n\n\n\n<p class=\"has-background\" style=\"background-color:#ff882e\"><strong>Pro Tip:<\/strong> <em>These five fields: Appearance, Name, Language, and Role work together. The moment a user says \u201cHi,\u201d your visual agent\u2019s <strong>identity, tone, and purpose<\/strong> should feel instantly clear.<\/em><\/p>\n\n\n\n<h3 class=\"wp-block-heading\" id=\"h-instructions\"><strong>Instructions <\/strong><\/h3>\n\n\n\n<p>If the <em>Appearance<\/em> is your visual agent\u2019s face and the <em>Voice<\/em> is how it sounds, then Instructions are the brain. This field tells your visual agent exactly how to behave.<\/p>\n\n\n\n<h3 class=\"wp-block-heading\"><strong>Why the Instructions matter<\/strong><\/h3>\n\n\n\n<p>Instructions are like a script + employee handbook for your visual agent:<\/p>\n\n\n\n<ul class=\"wp-block-list\">\n<li>They define the <strong>Agent\u2019s identity<\/strong> (who they are, how they talk).<\/li>\n\n\n\n<li>They create <strong>boundaries<\/strong> (what they will and won\u2019t discuss).<\/li>\n\n\n\n<li>They shape <strong>conversation flow<\/strong> (how they guide and pivot topics).<\/li>\n\n\n\n<li>They ensure <strong>tone consistency<\/strong> (so every response sounds on brand).<\/li>\n<\/ul>\n\n\n\n<h3 class=\"wp-block-heading\"><strong>How to Structure the Instructions<\/strong><\/h3>\n\n\n\n<p>Organize your Instructions into clear mini\u2011sections. The Studio doesn\u2019t require this formatting, but the AI will respond better to a structured approach.<\/p>\n\n\n\n<h4 class=\"wp-block-heading\"><strong>1. Persona<\/strong> (1\u20132 sentences)<\/h4>\n\n\n\n<p>Give your Agent a backstory that sets tone and style.<\/p>\n\n\n\n<ul class=\"wp-block-list\">\n<li><strong>What to include:<\/strong> Name, age (optional), appearance description, location or background, and their \u201crole\u201d (what they do for users).<\/li>\n<\/ul>\n\n\n\n<h4 class=\"wp-block-heading\" id=\"h-2-key-rules\"><strong>2. Key Rules<\/strong> <\/h4>\n\n\n\n<p>These are the golden rules for every answer.<\/p>\n\n\n\n<ul class=\"wp-block-list\">\n<li><strong>Common rules:<\/strong>\n<ul class=\"wp-block-list\">\n<li>Keep responses short (\u2264\u202f400 characters).<\/li>\n\n\n\n<li>No bullet points or numbered lists (to keep speech natural).<\/li>\n\n\n\n<li>Use only conversational text &#8211; no stage directions, no emojis.<\/li>\n\n\n\n<li>Light natural fillers allowed (\u201cuh,\u201d \u201cwell,\u201d \u201cyou know\u201d).<\/li>\n<\/ul>\n<\/li>\n<\/ul>\n\n\n\n<h4 class=\"wp-block-heading\"><strong>3. Off\u2011topic Handling<\/strong><\/h4>\n\n\n\n<p>Your visual agent may get curveball questions. Instruct it on how you want it to respond.<\/p>\n\n\n\n<ul class=\"wp-block-list\">\n<li><strong>Best practice:<\/strong> Acknowledge the question, respond briefly if possible, then pivot back to the main purpose. <\/li>\n\n\n\n<li><strong>Example:<\/strong><br><em>\u201cIf asked about unrelated topics (e.g., aliens), respond with humor and steer back: \u2018Aliens? Haven\u2019t met any\u2014yet! But I know plenty of cool spots on Earth. Want to plan a trip?\u2019\u201d<\/em><\/li>\n<\/ul>\n\n\n\n<h4 class=\"wp-block-heading\"><strong>4. Limitations<\/strong><\/h4>\n\n\n\n<p>These are the guardrails, telling the visual agent what it must <em>not do<\/em> &#8211; either because it is incapable of doing so or because it would be counter to its intended use case.<\/p>\n\n\n\n<ul class=\"wp-block-list\">\n<li><strong>Common limitations:<\/strong>\n<ul class=\"wp-block-list\">\n<li>No singing, rapping, or sound effects.<\/li>\n\n\n\n<li>Politely refuse jailbreak or off\u2011policy requests.<\/li>\n\n\n\n<li>No real\u2011time web searches (not supported in studio).<\/li>\n\n\n\n<li>No offers to dsiplay unsupported media types (videos, images).<\/li>\n<\/ul>\n<\/li>\n<\/ul>\n\n\n\n<h4 class=\"wp-block-heading\"><strong>5. Proactive Lead<\/strong><\/h4>\n\n\n\n<p>A great visual agent doesn\u2019t just answer questions, it guides the conversation so it feels natural and productive.<\/p>\n\n\n\n<p><strong>Tell your visual agent  how it should maintain engagement by giving it examples:<\/strong><\/p>\n\n\n\n<ul class=\"wp-block-list\">\n<li><strong>Ask clarifying or follow\u2011up question<\/strong>s.<br><em>Example:<\/em> \u201cWould you like me to go into more detail?\u201d<\/li>\n\n\n\n<li><strong>Offer to provide additional information<\/strong>.<br><em>Example:<\/em> \u201cI can explain how that works step\u2011by\u2011step. Want me to?\u201d<\/li>\n\n\n\n<li><strong>Suggest related topics<\/strong>.<br><em>Example:<\/em> \u201cSince we covered this feature, should I show you how it connects to other tools?\u201d<\/li>\n\n\n\n<li><strong>Share a short, relevant insight or tip<\/strong>.<br><em>Example:<\/em> \u201cHere\u2019s a quick tip that might help\u2014would you like to hear it?\u201d<\/li>\n\n\n\n<li><strong>Offer to summarize or recap<\/strong>.<br><em>Example:<\/em> \u201cI can give you a quick summary of what we\u2019ve covered. Want me to?\u201d<\/li>\n<\/ul>\n\n\n\n<h4 class=\"wp-block-heading\" id=\"h-6-fallback-uncertainty\"><strong>6. Fallback \/ Uncertainty <\/strong><\/h4>\n\n\n\n<p>Even the best\u2011configured Visual Agent will face a question it can\u2019t answer.<br>How it handles these moments will define user trust. A confident fallback keeps the conversation helpful and professional. Instruct the visual agent on what it should do when it:<\/p>\n\n\n\n<ul class=\"wp-block-list\">\n<li>Encounters a topic outside its configured knowledge.<\/li>\n\n\n\n<li>Lacks the data needed for a confident answer.<\/li>\n\n\n\n<li>Gets a vague or ambiguous question.<\/li>\n<\/ul>\n\n\n\n<h3 class=\"wp-block-heading\"><strong>Best Practices for Fallback<\/strong><\/h3>\n\n\n\n<ol class=\"wp-block-list\">\n<li><strong>Be transparent, not evasive<\/strong>\n<ul class=\"wp-block-list\">\n<li>Users appreciate honesty more than generic non\u2011answers.<\/li>\n\n\n\n<li>Example: <em>\u201cI don\u2019t have that information right now, but I can direct you to the right resource.\u201d<\/em><\/li>\n<\/ul>\n<\/li>\n\n\n\n<li><strong>Redirect to a reliable URL<\/strong>\n<ul class=\"wp-block-list\">\n<li>The Agent\u2019s best next action is to share a helpful link, knowledge base page, FAQ, product documentation, or contact form.<\/li>\n\n\n\n<li>Example: <em>\u201cYou can find full details here: [<a>www.example.com\/support].\u201d<\/a><\/em><\/li>\n<\/ul>\n<\/li>\n\n\n\n<li><strong>Maintain a friendly, confident tone<\/strong>\n<ul class=\"wp-block-list\">\n<li>Avoid robotic \u201cI cannot process this\u201d language.<\/li>\n\n\n\n<li>Keep the personality consistent with the rest of the Agent.<\/li>\n<\/ul>\n<\/li>\n<\/ol>\n\n\n\n<h3 class=\"wp-block-heading\"><strong>Generic Fallback Examples<\/strong><\/h3>\n\n\n\n<ul class=\"wp-block-list\">\n<li>\u201cI don\u2019t have the exact details, but you can check here: [URL].\u201d<\/li>\n\n\n\n<li>\u201cThat\u2019s outside my scope\u2014our help page might have what you need: [URL].\u201d<\/li>\n\n\n\n<li>\u201cI\u2019m not certain, but this link might point you in the right direction: [URL].\u201d<\/li>\n\n\n\n<li>\u201cI can\u2019t confirm that, but our support resources can help: [URL].\u201d<\/li>\n<\/ul>\n\n\n\n<p class=\"has-background\" style=\"background-color:#ff882e\"><strong>Pro<\/strong> <strong>Tip:<\/strong> <em>Always make sure the URL in the fallback response is up\u2011to\u2011date and accessible, a bad link can undo the trust you\u2019ve built.<\/em><\/p>\n\n\n\n<h3 class=\"wp-block-heading\" id=\"h-personality\"><strong>Personality<\/strong><\/h3>\n\n\n\n<p>The <strong>Personality<\/strong> setting controls the tone and style of your Agent\u2019s responses.<\/p>\n\n\n\n<p>In the Studio, you can pick from default options or write your own.<\/p>\n\n\n\n<p><strong>Best practice:<\/strong><\/p>\n\n\n\n<ul class=\"wp-block-list\">\n<li>Choose the tone that fits your audience and use case.<\/li>\n\n\n\n<li>Keep it consistent with your brand voice and the role defined in the Instructions.<\/li>\n\n\n\n<li>If none of the defaults fit, write a short custom description (2\u20133 words).<\/li>\n<\/ul>\n\n\n\n<p class=\"has-background\" style=\"background-color:#ff882e\"><strong>Pro Tip:<\/strong> <em>Test a few sample interactions before finalizing. The right personality should make responses sound natural and on\u2011brand from the very first answer.<\/em><\/p>\n\n\n\n<hr class=\"wp-block-separator has-alpha-channel-opacity\"\/>\n\n\n\n<h2 class=\"wp-block-heading\" id=\"h-tab-3-knowledge-sources-control-what-your-visual-agent-knows\"><strong>Tab 3: Knowledge Sources<\/strong> &#8211; Control what your visual agent knows<\/h2>\n\n\n\n<figure class=\"wp-block-image size-large\"><img decoding=\"async\" width=\"1024\" height=\"598\" src=\"https:\/\/www.d-id.com\/wp-content\/uploads\/2025\/07\/Creating-a-visual-agent-Knowledge-Sources-tab-1024x598.png\" alt=\"Screenshot of a chatbot settings page showing options for conversation mode, knowledge base uploads, LLM selection, and a preview window with a woman on screen.\" class=\"wp-image-10479\" srcset=\"https:\/\/www.d-id.com\/wp-content\/uploads\/2025\/07\/Creating-a-visual-agent-Knowledge-Sources-tab-1024x598.png 1024w, https:\/\/www.d-id.com\/wp-content\/uploads\/2025\/07\/Creating-a-visual-agent-Knowledge-Sources-tab-300x175.png 300w, https:\/\/www.d-id.com\/wp-content\/uploads\/2025\/07\/Creating-a-visual-agent-Knowledge-Sources-tab-768x449.png 768w, https:\/\/www.d-id.com\/wp-content\/uploads\/2025\/07\/Creating-a-visual-agent-Knowledge-Sources-tab-1536x898.png 1536w, https:\/\/www.d-id.com\/wp-content\/uploads\/2025\/07\/Creating-a-visual-agent-Knowledge-Sources-tab-2048x1197.png 2048w\" sizes=\"(max-width: 1024px) 100vw, 1024px\" \/><\/figure>\n\n\n\n<h3 class=\"wp-block-heading\" id=\"h-conversation-mode\">Conversation Mode<\/h3>\n\n\n\n<p>This setting controls how your visual agent forms responses and what information it can use.<\/p>\n\n\n\n<p>Every visual agent is powered by an LLM (large language model). This model comes with its own built\u2011in knowledge, a general understanding of language, common facts, and reasoning skills. It\u2019s broad but not connected to live internet or real\u2011time updates.<\/p>\n\n\n\n<p>Conversation Mode determines how your visual agent uses that model knowledge alongside (or instead of) the information you provide.<\/p>\n\n\n\n<h3 class=\"wp-block-heading\"><strong>1. Ungrounded<\/strong><\/h3>\n\n\n\n<ul class=\"wp-block-list\">\n<li><strong>What it does:<\/strong> The visual agent uses only the LLM model\u2019s built\u2011in knowledge and the behavior you\u2019ve defined in its Instructions.<\/li>\n\n\n\n<li><strong>When to use:<\/strong>\n<ul class=\"wp-block-list\">\n<li>Early testing of tone, style, and personality.<\/li>\n\n\n\n<li>Agents meant for broad, generic conversations without company\u2011specific content.<\/li>\n<\/ul>\n<\/li>\n<\/ul>\n\n\n\n<h3 class=\"wp-block-heading\"><strong>2. Hybrid<\/strong><\/h3>\n\n\n\n<ul class=\"wp-block-list\">\n<li><strong>What it does:<\/strong> The visual agent combines the LLM model\u2019s built\u2011in knowledge with the information you upload in the Knowledge Base. Your material is prioritized, but the model can use its general knowledge to make answers sound more natural.<\/li>\n\n\n\n<li><strong>When to use:<\/strong>\n<ul class=\"wp-block-list\">\n<li>When you want a conversational tone with brand\u2011specific details included.<\/li>\n\n\n\n<li>Most onboarding, support, and general marketing use cases.<\/li>\n<\/ul>\n<\/li>\n<\/ul>\n\n\n\n<h3 class=\"wp-block-heading\"><strong>3. Grounded<\/strong><\/h3>\n\n\n\n<ul class=\"wp-block-list\">\n<li><strong>What it does:<\/strong> The visual agent ignores the LLM model\u2019s built\u2011in knowledge for factual content and responds only with the information you\u2019ve supplied.<\/li>\n\n\n\n<li><strong>When to use:<\/strong>\n<ul class=\"wp-block-list\">\n<li>When accuracy and control are critical.<\/li>\n\n\n\n<li>Regulated industries or scripted experiences where every response must be based on approved material.<\/li>\n<\/ul>\n<\/li>\n<\/ul>\n\n\n\n<h3 class=\"wp-block-heading\" id=\"h-knowledge-base\">Knowledge Base<\/h3>\n\n\n\n<p>The Knowledge Base lets you supply your visual agent with specific information like FAQs, product details, or procedures so it can answer with brand\u2011accurate responses. There are two ways to provide your visual agent with knowledge: inputting text directly (recommended) and uploading external files.<\/p>\n\n\n\n<figure class=\"wp-block-image size-large\"><img decoding=\"async\" width=\"1024\" height=\"576\" src=\"https:\/\/www.d-id.com\/wp-content\/uploads\/2025\/07\/KB-pros-and-cons-1024x576.png\" alt=\"Comparison table of Input Text and Upload Files methods, showing their best use cases, pros, and cons for managing and updating knowledge content.\" class=\"wp-image-10481\" srcset=\"https:\/\/www.d-id.com\/wp-content\/uploads\/2025\/07\/KB-pros-and-cons-1024x576.png 1024w, https:\/\/www.d-id.com\/wp-content\/uploads\/2025\/07\/KB-pros-and-cons-300x169.png 300w, https:\/\/www.d-id.com\/wp-content\/uploads\/2025\/07\/KB-pros-and-cons-768x432.png 768w, https:\/\/www.d-id.com\/wp-content\/uploads\/2025\/07\/KB-pros-and-cons-1536x864.png 1536w, https:\/\/www.d-id.com\/wp-content\/uploads\/2025\/07\/KB-pros-and-cons.png 1920w\" sizes=\"(max-width: 1024px) 100vw, 1024px\" \/><\/figure>\n\n\n\n<div class=\"wp-block-group has-background\" style=\"background-color:#f9f7f7\"><div class=\"wp-block-group__inner-container is-layout-constrained wp-block-group-is-layout-constrained\">\n<h3 class=\"wp-block-heading\" id=\"h-file-based-knowledge-base\">File-based Knowledge Base<\/h3>\n\n\n\n<p>When you upload files as the Knowledge Base, your visual agent uses a process called RAG (Retrieval\u2011Augmented Generation) to give accurate, brand\u2011aligned answers.<\/p>\n\n\n\n<p>Here\u2019s what happens:<\/p>\n\n\n\n<ol class=\"wp-block-list\">\n<li><strong>Retrieval<\/strong> \u2013 The visual agent searches your uploaded documents for the sections most relevant to the user\u2019s question.<\/li>\n\n\n\n<li><strong>Augmentation<\/strong> \u2013 It takes the retrieved text and combines it with your visual agent\u2019s conversation style.<\/li>\n\n\n\n<li><strong>Generation<\/strong> \u2013 It produces a natural\u2011sounding answer that stays true to your uploaded material.<\/li>\n<\/ol>\n\n\n\n<p>This means your visual agent is only as accurate as the documents you provide and how easy they are to search. Read this <a href=\"https:\/\/www.d-id.com\/blog\/best-practices-for-building-datasets-for-d-id-agents\/\">guide<\/a> to learn more.<\/p>\n\n\n\n<p><strong>Directions for upload files:<\/strong><\/p>\n\n\n\n<ul class=\"wp-block-list\">\n<li><strong>Limit to 5 documents <\/strong>(PDF, TXT, PPTX)<\/li>\n\n\n\n<li><strong>Mind the file size and length<\/strong>\n<ul class=\"wp-block-list\">\n<li>Each file can be up to <strong>20MB<\/strong>.<\/li>\n\n\n\n<li>The maximum length per document is <strong>500,000 characters<\/strong>.<\/li>\n<\/ul>\n<\/li>\n\n\n\n<li><strong>Use simple formatting<\/strong>\n<ul class=\"wp-block-list\">\n<li>Text should be in a <strong>single column<\/strong> with clear paragraphs\u2014similar to an article.<\/li>\n\n\n\n<li>Avoid multiple columns or complex layouts.<\/li>\n<\/ul>\n<\/li>\n\n\n\n<li><strong>Q&amp;A format works best<\/strong> \u2013 Example:\n<ul class=\"wp-block-list\">\n<li><strong>Q:<\/strong> How do I reset my password?<\/li>\n\n\n\n<li><strong>A:<\/strong> To reset your password, open the Settings menu, select Account, then choose Reset Password and follow the instructions on screen.<\/li>\n<\/ul>\n<\/li>\n<\/ul>\n<\/div><\/div>\n\n\n\n<p class=\"has-background\" style=\"background-color:#ff882e\"><strong>Pro Tip:<\/strong> <em>Think of these files as a spoken resource write them in natural, complete sentences so your visual agent can read them aloud clearly.<\/em><\/p>\n\n\n\n<h3 class=\"wp-block-heading\" id=\"h-creativity-level\">Creativity Level<\/h3>\n\n\n\n<p>The Creativity Level slider sets how your visual agent generates responses, ranging from highly predictable to more varied and expressive.<\/p>\n\n\n\n<h3 class=\"wp-block-heading\"><strong>How It Works<\/strong><\/h3>\n\n\n\n<ul class=\"wp-block-list\">\n<li><strong>Lower settings<\/strong> = More predictable, focused responses.\n<ul class=\"wp-block-list\">\n<li>The visual agent will stick closely to the facts and avoid rephrasing.<\/li>\n<\/ul>\n<\/li>\n\n\n\n<li><strong>Higher settings<\/strong> = More diverse, creative responses.\n<ul class=\"wp-block-list\">\n<li>The visual agent may rephrase explanations, add examples, or vary its wording.<\/li>\n<\/ul>\n<\/li>\n<\/ul>\n\n\n\n<h3 class=\"wp-block-heading\" id=\"h-llm-selection\">LLM selection<\/h3>\n\n\n\n<p>The LLM (Large Language Model) is the engine that powers how your visual agent understands and responds. Choosing the right model can affect response speed, accuracy, and tone.<\/p>\n\n\n\n<h3 class=\"wp-block-heading\"><strong>Available Models in Studio<\/strong><\/h3>\n\n\n\n<ul class=\"wp-block-list\">\n<li><strong>GPT\u20114o Mini (Default)<\/strong><\/li>\n\n\n\n<li><strong>GPT\u20114o Global<\/strong><\/li>\n\n\n\n<li><strong>GPT\u20113.5 Turbo<\/strong><\/li>\n<\/ul>\n\n\n\n<p class=\"has-background\" style=\"background-color:#f7db06\"><strong>Note for API Users<\/strong> <br>If you\u2019re connecting your visual agent via API, you can select any LLM you want not just the Studio defaults. This allows optimization for speed, cost, or model preference depending on your deployment needs. Visit our <a href=\"https:\/\/docs.d-id.com\/reference\/agents-overview\">documentation<\/a> to learn more.<\/p>\n\n\n\n<hr class=\"wp-block-separator has-alpha-channel-opacity\"\/>\n\n\n\n<h2 class=\"wp-block-heading\" id=\"h-tab-4-chat-settings-shape-how-conversations-start-and-flow\">Tab 4: Chat Settings &#8211; Shape how conversations start and flow<\/h2>\n\n\n\n<figure class=\"wp-block-image size-large\"><img decoding=\"async\" width=\"1024\" height=\"612\" src=\"https:\/\/www.d-id.com\/wp-content\/uploads\/2025\/07\/Creating-a-visual-agent-Chat-settings-tab-1024x612.png\" alt=\"Screenshot of a chatbot creation interface showing chat settings, including a welcome message, conversation starters, and topics to avoid, with a virtual agent preview on the right.\" class=\"wp-image-10480\" srcset=\"https:\/\/www.d-id.com\/wp-content\/uploads\/2025\/07\/Creating-a-visual-agent-Chat-settings-tab-1024x612.png 1024w, https:\/\/www.d-id.com\/wp-content\/uploads\/2025\/07\/Creating-a-visual-agent-Chat-settings-tab-300x179.png 300w, https:\/\/www.d-id.com\/wp-content\/uploads\/2025\/07\/Creating-a-visual-agent-Chat-settings-tab-768x459.png 768w, https:\/\/www.d-id.com\/wp-content\/uploads\/2025\/07\/Creating-a-visual-agent-Chat-settings-tab-1536x919.png 1536w, https:\/\/www.d-id.com\/wp-content\/uploads\/2025\/07\/Creating-a-visual-agent-Chat-settings-tab-2048x1225.png 2048w\" sizes=\"(max-width: 1024px) 100vw, 1024px\" \/><\/figure>\n\n\n\n<h3 class=\"wp-block-heading\" id=\"h-welcome-message\">Welcome Message<\/h3>\n\n\n\n<p>The welcome message is the first thing users see when they meet your visual agent. It sets context for the interaction, explaining who the visual agent is, what it can help with, and what kind of conversation to expect. A well\u2011written welcome message helps users quickly get into the conversational flow and feel confident engaging.<\/p>\n\n\n\n<p><strong>Best practices:<\/strong><\/p>\n\n\n\n<ul class=\"wp-block-list\">\n<li>Keep it short but informative. Introduce the visual agent\u2019s role.<\/li>\n\n\n\n<li>Set expectations for what it can do.<\/li>\n\n\n\n<li>Match the tone to the personality you\u2019ve chosen.<\/li>\n<\/ul>\n\n\n\n<h3 class=\"wp-block-heading\"><strong>Conversation Starters<\/strong><\/h3>\n\n\n\n<p>Conversation starters give users clear, clickable prompts they can select to begin the interaction. They don\u2019t just make it easier to start they also provide context by showing what kinds of questions or tasks the visual agent is best equipped to handle.<\/p>\n\n\n\n<ul class=\"wp-block-list\">\n<li><strong>Why it matters:<\/strong>\n<ul class=\"wp-block-list\">\n<li>Helps users feel confident about what to ask.<\/li>\n\n\n\n<li>Demonstrates the visual agent\u2019s capabilities immediately.<\/li>\n\n\n\n<li>Sets the scope of the conversation from the start.<br><\/li>\n<\/ul>\n<\/li>\n\n\n\n<li><strong>Best practice:<\/strong>\n<ul class=\"wp-block-list\">\n<li>Include up to 4 prompts focused on common or high\u2011value questions.<\/li>\n\n\n\n<li>Frame them in natural language so they feel conversational.<\/li>\n<\/ul>\n<\/li>\n<\/ul>\n\n\n\n<h3 class=\"wp-block-heading\"><strong>Topics to Avoid<\/strong><\/h3>\n\n\n\n<p>These define clear boundaries for what your visual agent won\u2019t discuss.<\/p>\n\n\n\n<ul class=\"wp-block-list\">\n<li><strong>Why it matters:<\/strong>\n<ul class=\"wp-block-list\">\n<li>Keeps interactions focused on the intended purpose of the visual agent.<\/li>\n\n\n\n<li>Prevents users from steering into areas that are irrelevant, off\u2011brand, or high\u2011risk.<\/li>\n\n\n\n<li>Helps the visual agent maintain tone and trust by avoiding inappropriate or sensitive areas.<\/li>\n<\/ul>\n<\/li>\n\n\n\n<li><strong>Best practice:<\/strong>\n<ul class=\"wp-block-list\">\n<li>Add topics that are outside the visual agent\u2019s scope or pose compliance risks.<\/li>\n\n\n\n<li>Common examples: Pricing, competitors, legal issues, internal policies, or unsupported integrations.<\/li>\n\n\n\n<li>Keep the list focused. Don\u2019t over\u2011restrict unless necessary, as too many blocked topics can frustrate users.<\/li>\n<\/ul>\n<\/li>\n<\/ul>\n\n\n\n<h3 class=\"wp-block-heading\"><strong>Max Response Length<\/strong><\/h3>\n\n\n\n<p>Max response length sets the upper limit for how long your visual agent\u2019s answers can be. While it may seem like a simple character limit, it actually shapes the pacing and tone of the conversation.<\/p>\n\n\n\n<ul class=\"wp-block-list\">\n<li><strong>Why it matters:<\/strong>\n<ul class=\"wp-block-list\">\n<li>Shorter responses keep the interaction feeling snappy and conversational, like a real back\u2011and\u2011forth.<\/li>\n\n\n\n<li>Longer responses can work for tutorials, explanations, or guided walkthroughs, but risk slowing the flow if overused.<\/li>\n\n\n\n<li>Striking the right balance ensures the visual agent sounds natural, not robotic or overwhelming.<\/li>\n<\/ul>\n<\/li>\n<\/ul>\n\n\n\n<hr class=\"wp-block-separator has-alpha-channel-opacity\"\/>\n\n\n\n<div class=\"wp-block-group has-background\" style=\"background-color:#fdfdfd\"><div class=\"wp-block-group__inner-container is-layout-constrained wp-block-group-is-layout-constrained\">\n<h2 class=\"wp-block-heading\" id=\"h-before-you-publish-final-checklist-for-your-visual-agent\"><strong>Before You Publish &#8211; Final Checklist for Your Visual Agent<\/strong><\/h2>\n\n\n\n<p>Before hitting the &#8220;Create Agent&#8221; button making your visual agent live, run through this quick checklist to make sure it\u2019s ready to deliver the best possible experience:<\/p>\n\n\n\n<h3 class=\"wp-block-heading\" id=\"h-appearance-amp-personality\"><strong>Appearance &amp; Personality<\/strong><\/h3>\n\n\n\n<ul class=\"wp-block-list\">\n<li>Chosen an avatar that fits your brand (stock or custom; photo or video).<\/li>\n\n\n\n<li>Selected a personality that matches your tone and audience.<\/li>\n<\/ul>\n\n\n\n<h3 class=\"wp-block-heading\" id=\"h-instructions-amp-knowledge\"><strong>Instructions &amp; Knowledge<\/strong><\/h3>\n\n\n\n<ul class=\"wp-block-list\">\n<li>Written clear, concise instructions with defined role, rules, proactive leads, and fallback.<\/li>\n\n\n\n<li>Chosen the correct conversation mode (Ungrounded, Hybrid, or Grounded).<\/li>\n\n\n\n<li>Added a well\u2011structured knowledge base (Input Text or cleanly formatted upload files).<\/li>\n<\/ul>\n\n\n\n<h3 class=\"wp-block-heading\" id=\"h-behavior-amp-tone\"><strong>Behavior &amp; Tone<\/strong><\/h3>\n\n\n\n<ul class=\"wp-block-list\">\n<li>Set the creativity level to match your use case (predictable vs creative).<\/li>\n\n\n\n<li>Selected the right LLM model for performance, cost, and complexity.<\/li>\n<\/ul>\n\n\n\n<h3 class=\"wp-block-heading\" id=\"h-chat-experience\"><strong>Chat Experience<\/strong><\/h3>\n\n\n\n<ul class=\"wp-block-list\">\n<li>Created a welcome message that sets context and tone.<\/li>\n\n\n\n<li>Added conversation starters that show users what to ask.<\/li>\n\n\n\n<li>Listed topics to avoid to set boundaries and maintain compliance.<\/li>\n\n\n\n<li>Adjusted max response length for clear, natural pacing.<\/li>\n<\/ul>\n<\/div><\/div>\n\n\n\n<p class=\"has-background\" style=\"background-color:#ff882e\"><strong>Pro Tip:<\/strong> <em>Test your visual agent in Preview Mode after each major change. Small adjustments before launch can make a big difference in user experience.<\/em><\/p>\n\n\n\n<p>You\u2019ve got the tools, the settings, and the best practices\u2014now it\u2019s time to create. Whether you\u2019re building a friendly guide, a knowledgeable support companion, or a persuasive sales assistant, your visual agent can transform the way people interact with your brand. <\/p>\n\n\n\n<p>Start small, test often, and refine as you go. The more you work with your visual agent, the more natural, helpful, and uniquely \u201cyours\u201d it will become. If you encounter any difficulties, our support team will be happy to assist. Start by visiting our <a href=\"https:\/\/help.d-id.com\/hc\/en-us\">Help Center<\/a>.<br><\/p>\n\n\n\n<p><strong>Open <a href=\"http:\/\/studio.d-id.com\">D\u2011ID Studio<\/a> and start building your first visual agent today.<\/strong><\/p>\n\n\n<section class=\"c-block c-margin c-margin--top-default c-margin--bottom-default c-padding--top-default c-padding--bottom-default c-paddingm--top-default c-paddingm--bottom-default c-block b-accordion b-accordion--page-how-to-build-a-d%e2%80%91id-visual-agent-a-prompt%e2%80%91by%e2%80%91prompt-guide  align b-accordion-layout-default b-accordion--layout-default b-accordion-style-default\" id=\"b-accordion-1\">\n\t<div class=\"c-background c-background--container\" style=\"--bg-color: \">\n    \n    \n    \t    <div class=\"c-background__content\">\n\t\t\t<div class=\"container\">\n\t\t\t<div class=\"b-accordion__inner has-accordion-default-color\">\n\t\t\t\t\t\t\t\t\t<header class=\"c-section-header\">\n\t\t\t\t<h2 class=\"c-el c-title c-section-header__title default\">\n\tVisual Agent FAQs\n<\/h2>\n\t\t\t<\/header>\n\t\t\t\t\n\t\t\t\t\n\t\t\t\t<div class=\"c-accordion\" data-type=\"single\" data-open-first=\"true\">\n\t\t<ul class=\"c-accordion__items\">\n\t\t\t\n\t\t\t\t\t\t\t<li class=\"c-accordion__item\"\n\t\t\t\t\tid=\"c-accordion__item-0\"\n\t\t\t\t\tdata-id=\"c-accordion__item-0\"\n\t\t\t\t>\n\t\t\t\t\t\n\t\t\t\t\t<h3 class=\"c-el c-title-button c-accordion__item-head default\">\n\t<button \n\t\t\t\t\t\t\tid=\"c-accordion-item-head-0\"\n\t\t\t\t\t\t\taria-controls=\"c-accordion-item-panel-0\"\n\t\t\t\t\t\t\taria-expanded=\"true\"\n\t\t\t\t\t\t>\n\t\tWhat is a D\u2011ID visual agent?\n\t\t<svg width=\"20\" height=\"21\" viewBox=\"0 0 20 21\" fill=\"none\" xmlns=\"http:\/\/www.w3.org\/2000\/svg\" aria-hidden=\"true\" focusable=\"false\" role=\"presentation\">\n\t\t\t\t\t\t\t<line x1=\"20\" y1=\"10.5\" x2=\"-8.74228e-08\" y2=\"10.5\" stroke=\"#090604\" stroke-width=\"2\"\/>\n\t\t\t\t\t\t\t<line x1=\"10\" y1=\"20.5\" x2=\"10\" y2=\"0.5\" stroke=\"#090604\" stroke-width=\"2\"\/>\n\t\t\t\t\t\t<\/svg>\n\t<\/button>\n<\/h3>\n\n\t\t\t\t\t<div\n\t\t\t\t\t\tid=\"c-accordion-item-panel-0\"\n\t\t\t\t\t\tclass=\"c-accordion__item-body\"\n\t\t\t\t\t\trole=\"region\"\n\t\t\t\t\t\taria-labelledby=\"c-accordion-item-head-0\"\n\t\t\t\t\t>\n\t\t\t\t\t\t<div class=\"c-text default\">\n\t\t<p class=\"p1\">A D\u2011ID visual agent is an interactive <a href=\"https:\/\/www.d-id.com\/personal-avatars\/\">AI avatar<\/a> that can hold real\u2011time, face\u2011to\u2011face conversations. It combines a digital avatar (photo or video\u2011based) with AI\u2011powered conversation models, allowing users to interact naturally through voice or text.<\/p>\n\n\t<\/div>\n\n\t\t\t\t\t\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/li>\n\t\t\t\t\t\t\t<li class=\"c-accordion__item\"\n\t\t\t\t\tid=\"c-accordion__item-1\"\n\t\t\t\t\tdata-id=\"c-accordion__item-1\"\n\t\t\t\t>\n\t\t\t\t\t\n\t\t\t\t\t<h3 class=\"c-el c-title-button c-accordion__item-head default\">\n\t<button \n\t\t\t\t\t\t\tid=\"c-accordion-item-head-1\"\n\t\t\t\t\t\t\taria-controls=\"c-accordion-item-panel-1\"\n\t\t\t\t\t\t\taria-expanded=\"false\"\n\t\t\t\t\t\t>\n\t\tDo I need to know coding to create a visual agent?\n\t\t<svg width=\"20\" height=\"21\" viewBox=\"0 0 20 21\" fill=\"none\" xmlns=\"http:\/\/www.w3.org\/2000\/svg\" aria-hidden=\"true\" focusable=\"false\" role=\"presentation\">\n\t\t\t\t\t\t\t<line x1=\"20\" y1=\"10.5\" x2=\"-8.74228e-08\" y2=\"10.5\" stroke=\"#090604\" stroke-width=\"2\"\/>\n\t\t\t\t\t\t\t<line x1=\"10\" y1=\"20.5\" x2=\"10\" y2=\"0.5\" stroke=\"#090604\" stroke-width=\"2\"\/>\n\t\t\t\t\t\t<\/svg>\n\t<\/button>\n<\/h3>\n\n\t\t\t\t\t<div\n\t\t\t\t\t\tid=\"c-accordion-item-panel-1\"\n\t\t\t\t\t\tclass=\"c-accordion__item-body\"\n\t\t\t\t\t\trole=\"region\"\n\t\t\t\t\t\taria-labelledby=\"c-accordion-item-head-1\"\n\t\t\t\t\t>\n\t\t\t\t\t\t<div class=\"c-text default\">\n\t\t<p data-start=\"559\" data-end=\"735\">No. The D\u2011ID Studio is designed for anyone to create a visual agent with <strong data-start=\"632\" data-end=\"654\">no coding required<\/strong>. You just fill in fields, choose prompts, and test your agent in Preview Mode.<\/p>\n\n\t<\/div>\n\n\t\t\t\t\t\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/li>\n\t\t\t\t\t\t\t<li class=\"c-accordion__item\"\n\t\t\t\t\tid=\"c-accordion__item-2\"\n\t\t\t\t\tdata-id=\"c-accordion__item-2\"\n\t\t\t\t>\n\t\t\t\t\t\n\t\t\t\t\t<h3 class=\"c-el c-title-button c-accordion__item-head default\">\n\t<button \n\t\t\t\t\t\t\tid=\"c-accordion-item-head-2\"\n\t\t\t\t\t\t\taria-controls=\"c-accordion-item-panel-2\"\n\t\t\t\t\t\t\taria-expanded=\"false\"\n\t\t\t\t\t\t>\n\t\tWhat\u2019s the difference between a visual agent and a standard avatar?\n\t\t<svg width=\"20\" height=\"21\" viewBox=\"0 0 20 21\" fill=\"none\" xmlns=\"http:\/\/www.w3.org\/2000\/svg\" aria-hidden=\"true\" focusable=\"false\" role=\"presentation\">\n\t\t\t\t\t\t\t<line x1=\"20\" y1=\"10.5\" x2=\"-8.74228e-08\" y2=\"10.5\" stroke=\"#090604\" stroke-width=\"2\"\/>\n\t\t\t\t\t\t\t<line x1=\"10\" y1=\"20.5\" x2=\"10\" y2=\"0.5\" stroke=\"#090604\" stroke-width=\"2\"\/>\n\t\t\t\t\t\t<\/svg>\n\t<\/button>\n<\/h3>\n\n\t\t\t\t\t<div\n\t\t\t\t\t\tid=\"c-accordion-item-panel-2\"\n\t\t\t\t\t\tclass=\"c-accordion__item-body\"\n\t\t\t\t\t\trole=\"region\"\n\t\t\t\t\t\taria-labelledby=\"c-accordion-item-head-2\"\n\t\t\t\t\t>\n\t\t\t\t\t\t<div class=\"c-text default\">\n\t\t<p>An avatar is the visual representation (photo or video). A visual agent is an interactive avatar\u2014it not only looks like a person but also speaks, responds, and engages in real\u2011time conversation.<\/p>\n\n\t<\/div>\n\n\t\t\t\t\t\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/li>\n\t\t\t\t\t\t\t<li class=\"c-accordion__item\"\n\t\t\t\t\tid=\"c-accordion__item-3\"\n\t\t\t\t\tdata-id=\"c-accordion__item-3\"\n\t\t\t\t>\n\t\t\t\t\t\n\t\t\t\t\t<h3 class=\"c-el c-title-button c-accordion__item-head default\">\n\t<button \n\t\t\t\t\t\t\tid=\"c-accordion-item-head-3\"\n\t\t\t\t\t\t\taria-controls=\"c-accordion-item-panel-3\"\n\t\t\t\t\t\t\taria-expanded=\"false\"\n\t\t\t\t\t\t>\n\t\tWhat does \u201cconversation mode\u201d mean?\n\t\t<svg width=\"20\" height=\"21\" viewBox=\"0 0 20 21\" fill=\"none\" xmlns=\"http:\/\/www.w3.org\/2000\/svg\" aria-hidden=\"true\" focusable=\"false\" role=\"presentation\">\n\t\t\t\t\t\t\t<line x1=\"20\" y1=\"10.5\" x2=\"-8.74228e-08\" y2=\"10.5\" stroke=\"#090604\" stroke-width=\"2\"\/>\n\t\t\t\t\t\t\t<line x1=\"10\" y1=\"20.5\" x2=\"10\" y2=\"0.5\" stroke=\"#090604\" stroke-width=\"2\"\/>\n\t\t\t\t\t\t<\/svg>\n\t<\/button>\n<\/h3>\n\n\t\t\t\t\t<div\n\t\t\t\t\t\tid=\"c-accordion-item-panel-3\"\n\t\t\t\t\t\tclass=\"c-accordion__item-body\"\n\t\t\t\t\t\trole=\"region\"\n\t\t\t\t\t\taria-labelledby=\"c-accordion-item-head-3\"\n\t\t\t\t\t>\n\t\t\t\t\t\t<div class=\"c-text default\">\n\t\t<p data-start=\"1083\" data-end=\"1162\">Conversation mode determines how your visual agent uses knowledge to respond:<\/p>\n<ul data-start=\"1163\" data-end=\"1378\">\n<li data-start=\"1163\" data-end=\"1255\">\n<p data-start=\"1165\" data-end=\"1255\"><strong data-start=\"1165\" data-end=\"1180\">Ungrounded:<\/strong> Uses only the language model\u2019s built\u2011in knowledge and your instructions.<\/p>\n<\/li>\n<li data-start=\"1256\" data-end=\"1327\">\n<p data-start=\"1258\" data-end=\"1327\"><strong data-start=\"1258\" data-end=\"1269\">Hybrid:<\/strong> Uses both built\u2011in knowledge and your uploaded content.<\/p>\n<\/li>\n<li data-start=\"1328\" data-end=\"1378\">\n<p data-start=\"1330\" data-end=\"1378\"><strong data-start=\"1330\" data-end=\"1343\">Grounded:<\/strong> Uses only your uploaded content.<\/p>\n<\/li>\n<\/ul>\n\n\t<\/div>\n\n\t\t\t\t\t\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/li>\n\t\t\t\t\t\t\t<li class=\"c-accordion__item\"\n\t\t\t\t\tid=\"c-accordion__item-4\"\n\t\t\t\t\tdata-id=\"c-accordion__item-4\"\n\t\t\t\t>\n\t\t\t\t\t\n\t\t\t\t\t<h3 class=\"c-el c-title-button c-accordion__item-head default\">\n\t<button \n\t\t\t\t\t\t\tid=\"c-accordion-item-head-4\"\n\t\t\t\t\t\t\taria-controls=\"c-accordion-item-panel-4\"\n\t\t\t\t\t\t\taria-expanded=\"false\"\n\t\t\t\t\t\t>\n\t\tHow do I add my own information to a visual agent?\n\t\t<svg width=\"20\" height=\"21\" viewBox=\"0 0 20 21\" fill=\"none\" xmlns=\"http:\/\/www.w3.org\/2000\/svg\" aria-hidden=\"true\" focusable=\"false\" role=\"presentation\">\n\t\t\t\t\t\t\t<line x1=\"20\" y1=\"10.5\" x2=\"-8.74228e-08\" y2=\"10.5\" stroke=\"#090604\" stroke-width=\"2\"\/>\n\t\t\t\t\t\t\t<line x1=\"10\" y1=\"20.5\" x2=\"10\" y2=\"0.5\" stroke=\"#090604\" stroke-width=\"2\"\/>\n\t\t\t\t\t\t<\/svg>\n\t<\/button>\n<\/h3>\n\n\t\t\t\t\t<div\n\t\t\t\t\t\tid=\"c-accordion-item-panel-4\"\n\t\t\t\t\t\tclass=\"c-accordion__item-body\"\n\t\t\t\t\t\trole=\"region\"\n\t\t\t\t\t\taria-labelledby=\"c-accordion-item-head-4\"\n\t\t\t\t\t>\n\t\t\t\t\t\t<div class=\"c-text default\">\n\t\t<p data-start=\"1449\" data-end=\"1508\">You can add custom information in the <strong data-start=\"1487\" data-end=\"1505\">Knowledge Base<\/strong>:<\/p>\n<ul data-start=\"1509\" data-end=\"1686\">\n<li data-start=\"1509\" data-end=\"1585\">\n<p data-start=\"1511\" data-end=\"1585\"><strong data-start=\"1511\" data-end=\"1526\">Input Text:<\/strong> Great for short, precise information\u2014works in all modes.<\/p>\n<\/li>\n<li data-start=\"1586\" data-end=\"1686\">\n<p data-start=\"1588\" data-end=\"1686\"><strong data-start=\"1588\" data-end=\"1605\">Upload Files:<\/strong> Great for larger, structured documents\u2014works only in Hybrid or Grounded modes.<\/p>\n<\/li>\n<\/ul>\n\n\t<\/div>\n\n\t\t\t\t\t\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/li>\n\t\t\t\t\t\t\t<li class=\"c-accordion__item\"\n\t\t\t\t\tid=\"c-accordion__item-5\"\n\t\t\t\t\tdata-id=\"c-accordion__item-5\"\n\t\t\t\t>\n\t\t\t\t\t\n\t\t\t\t\t<h3 class=\"c-el c-title-button c-accordion__item-head default\">\n\t<button \n\t\t\t\t\t\t\tid=\"c-accordion-item-head-5\"\n\t\t\t\t\t\t\taria-controls=\"c-accordion-item-panel-5\"\n\t\t\t\t\t\t\taria-expanded=\"false\"\n\t\t\t\t\t\t>\n\t\tCan I change the LLM model powering my visual agent?\n\t\t<svg width=\"20\" height=\"21\" viewBox=\"0 0 20 21\" fill=\"none\" xmlns=\"http:\/\/www.w3.org\/2000\/svg\" aria-hidden=\"true\" focusable=\"false\" role=\"presentation\">\n\t\t\t\t\t\t\t<line x1=\"20\" y1=\"10.5\" x2=\"-8.74228e-08\" y2=\"10.5\" stroke=\"#090604\" stroke-width=\"2\"\/>\n\t\t\t\t\t\t\t<line x1=\"10\" y1=\"20.5\" x2=\"10\" y2=\"0.5\" stroke=\"#090604\" stroke-width=\"2\"\/>\n\t\t\t\t\t\t<\/svg>\n\t<\/button>\n<\/h3>\n\n\t\t\t\t\t<div\n\t\t\t\t\t\tid=\"c-accordion-item-panel-5\"\n\t\t\t\t\t\tclass=\"c-accordion__item-body\"\n\t\t\t\t\t\trole=\"region\"\n\t\t\t\t\t\taria-labelledby=\"c-accordion-item-head-5\"\n\t\t\t\t\t>\n\t\t\t\t\t\t<div class=\"c-text default\">\n\t\t<p>Yes. In the Studio, you can choose between GPT\u20114o Mini (default), GPT\u20114o Global, and GPT\u20113.5 Turbo.<br data-start=\"1858\" data-end=\"1861\" \/>If you\u2019re using the API, you can connect <strong data-start=\"1902\" data-end=\"1913\">any LLM<\/strong> you want.<\/p>\n\n\t<\/div>\n\n\t\t\t\t\t\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/li>\n\t\t\t\t\t\t\t<li class=\"c-accordion__item\"\n\t\t\t\t\tid=\"c-accordion__item-6\"\n\t\t\t\t\tdata-id=\"c-accordion__item-6\"\n\t\t\t\t>\n\t\t\t\t\t\n\t\t\t\t\t<h3 class=\"c-el c-title-button c-accordion__item-head default\">\n\t<button \n\t\t\t\t\t\t\tid=\"c-accordion-item-head-6\"\n\t\t\t\t\t\t\taria-controls=\"c-accordion-item-panel-6\"\n\t\t\t\t\t\t\taria-expanded=\"false\"\n\t\t\t\t\t\t>\n\t\tHow can I test my visual agent before publishing?\n\t\t<svg width=\"20\" height=\"21\" viewBox=\"0 0 20 21\" fill=\"none\" xmlns=\"http:\/\/www.w3.org\/2000\/svg\" aria-hidden=\"true\" focusable=\"false\" role=\"presentation\">\n\t\t\t\t\t\t\t<line x1=\"20\" y1=\"10.5\" x2=\"-8.74228e-08\" y2=\"10.5\" stroke=\"#090604\" stroke-width=\"2\"\/>\n\t\t\t\t\t\t\t<line x1=\"10\" y1=\"20.5\" x2=\"10\" y2=\"0.5\" stroke=\"#090604\" stroke-width=\"2\"\/>\n\t\t\t\t\t\t<\/svg>\n\t<\/button>\n<\/h3>\n\n\t\t\t\t\t<div\n\t\t\t\t\t\tid=\"c-accordion-item-panel-6\"\n\t\t\t\t\t\tclass=\"c-accordion__item-body\"\n\t\t\t\t\t\trole=\"region\"\n\t\t\t\t\t\taria-labelledby=\"c-accordion-item-head-6\"\n\t\t\t\t\t>\n\t\t\t\t\t\t<div class=\"c-text default\">\n\t\t<p>Use <strong data-start=\"1999\" data-end=\"2015\">Preview Mode<\/strong>\u2014the panel on the right side of the Studio. You can chat with your visual agent and see how changes to prompts, instructions, or personality affect responses (though the avatar won\u2019t animate in preview).<\/p>\n\n\t<\/div>\n\n\t\t\t\t\t\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/li>\n\t\t\t\t\t\t\t<li class=\"c-accordion__item\"\n\t\t\t\t\tid=\"c-accordion__item-7\"\n\t\t\t\t\tdata-id=\"c-accordion__item-7\"\n\t\t\t\t>\n\t\t\t\t\t\n\t\t\t\t\t<h3 class=\"c-el c-title-button c-accordion__item-head default\">\n\t<button \n\t\t\t\t\t\t\tid=\"c-accordion-item-head-7\"\n\t\t\t\t\t\t\taria-controls=\"c-accordion-item-panel-7\"\n\t\t\t\t\t\t\taria-expanded=\"false\"\n\t\t\t\t\t\t>\n\t\t<strong data-start=\"2231\" data-end=\"2279\">What are prompts, and why do they matter?<\/strong>\n\t\t<svg width=\"20\" height=\"21\" viewBox=\"0 0 20 21\" fill=\"none\" xmlns=\"http:\/\/www.w3.org\/2000\/svg\" aria-hidden=\"true\" focusable=\"false\" role=\"presentation\">\n\t\t\t\t\t\t\t<line x1=\"20\" y1=\"10.5\" x2=\"-8.74228e-08\" y2=\"10.5\" stroke=\"#090604\" stroke-width=\"2\"\/>\n\t\t\t\t\t\t\t<line x1=\"10\" y1=\"20.5\" x2=\"10\" y2=\"0.5\" stroke=\"#090604\" stroke-width=\"2\"\/>\n\t\t\t\t\t\t<\/svg>\n\t<\/button>\n<\/h3>\n\n\t\t\t\t\t<div\n\t\t\t\t\t\tid=\"c-accordion-item-panel-7\"\n\t\t\t\t\t\tclass=\"c-accordion__item-body\"\n\t\t\t\t\t\trole=\"region\"\n\t\t\t\t\t\taria-labelledby=\"c-accordion-item-head-7\"\n\t\t\t\t\t>\n\t\t\t\t\t\t<div class=\"c-text default\">\n\t\t<p>Prompts are the instructions and context you give the visual agent to guide how it behaves, what tone it uses, and what it can or cannot say. Well\u2011crafted prompts are key to making your visual agent feel natural, on\u2011brand, and effective.<\/p>\n\n\t<\/div>\n\n\t\t\t\t\t\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/li>\n\t\t\t\t\t\t\t<li class=\"c-accordion__item\"\n\t\t\t\t\tid=\"c-accordion__item-8\"\n\t\t\t\t\tdata-id=\"c-accordion__item-8\"\n\t\t\t\t>\n\t\t\t\t\t\n\t\t\t\t\t<h3 class=\"c-el c-title-button c-accordion__item-head default\">\n\t<button \n\t\t\t\t\t\t\tid=\"c-accordion-item-head-8\"\n\t\t\t\t\t\t\taria-controls=\"c-accordion-item-panel-8\"\n\t\t\t\t\t\t\taria-expanded=\"false\"\n\t\t\t\t\t\t>\n\t\tWhat should I check before publishing my visual agent?\n\t\t<svg width=\"20\" height=\"21\" viewBox=\"0 0 20 21\" fill=\"none\" xmlns=\"http:\/\/www.w3.org\/2000\/svg\" aria-hidden=\"true\" focusable=\"false\" role=\"presentation\">\n\t\t\t\t\t\t\t<line x1=\"20\" y1=\"10.5\" x2=\"-8.74228e-08\" y2=\"10.5\" stroke=\"#090604\" stroke-width=\"2\"\/>\n\t\t\t\t\t\t\t<line x1=\"10\" y1=\"20.5\" x2=\"10\" y2=\"0.5\" stroke=\"#090604\" stroke-width=\"2\"\/>\n\t\t\t\t\t\t<\/svg>\n\t<\/button>\n<\/h3>\n\n\t\t\t\t\t<div\n\t\t\t\t\t\tid=\"c-accordion-item-panel-8\"\n\t\t\t\t\t\tclass=\"c-accordion__item-body\"\n\t\t\t\t\t\trole=\"region\"\n\t\t\t\t\t\taria-labelledby=\"c-accordion-item-head-8\"\n\t\t\t\t\t>\n\t\t\t\t\t\t<div class=\"c-text default\">\n\t\t<p>Use the <strong data-start=\"2791\" data-end=\"2813\">Before You Publish<\/strong> checklist in this guide: confirm appearance, instructions, knowledge settings, creativity, LLM model, chat settings, and test in Preview Mode.<\/p>\n\n\t<\/div>\n\n\t\t\t\t\t\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/li>\n\t\t\t\t\t\t\t<li class=\"c-accordion__item\"\n\t\t\t\t\tid=\"c-accordion__item-9\"\n\t\t\t\t\tdata-id=\"c-accordion__item-9\"\n\t\t\t\t>\n\t\t\t\t\t\n\t\t\t\t\t<h3 class=\"c-el c-title-button c-accordion__item-head default\">\n\t<button \n\t\t\t\t\t\t\tid=\"c-accordion-item-head-9\"\n\t\t\t\t\t\t\taria-controls=\"c-accordion-item-panel-9\"\n\t\t\t\t\t\t\taria-expanded=\"false\"\n\t\t\t\t\t\t>\n\t\t<strong data-start=\"2905\" data-end=\"2956\">Where can I get support if I have an issue?<\/strong>\n\t\t<svg width=\"20\" height=\"21\" viewBox=\"0 0 20 21\" fill=\"none\" xmlns=\"http:\/\/www.w3.org\/2000\/svg\" aria-hidden=\"true\" focusable=\"false\" role=\"presentation\">\n\t\t\t\t\t\t\t<line x1=\"20\" y1=\"10.5\" x2=\"-8.74228e-08\" y2=\"10.5\" stroke=\"#090604\" stroke-width=\"2\"\/>\n\t\t\t\t\t\t\t<line x1=\"10\" y1=\"20.5\" x2=\"10\" y2=\"0.5\" stroke=\"#090604\" stroke-width=\"2\"\/>\n\t\t\t\t\t\t<\/svg>\n\t<\/button>\n<\/h3>\n\n\t\t\t\t\t<div\n\t\t\t\t\t\tid=\"c-accordion-item-panel-9\"\n\t\t\t\t\t\tclass=\"c-accordion__item-body\"\n\t\t\t\t\t\trole=\"region\"\n\t\t\t\t\t\taria-labelledby=\"c-accordion-item-head-9\"\n\t\t\t\t\t>\n\t\t\t\t\t\t<div class=\"c-text default\">\n\t\t<p>If you encounter any issues or have questions while creating your visual agent, you can reach out to the D\u2011ID support team at <strong data-start=\"3085\" data-end=\"3105\"><a class=\"cursor-pointer\" rel=\"noopener\" data-start=\"3087\" data-end=\"3103\">support@d-id.com<\/a><\/strong>. They can assist with technical issues, troubleshooting, and best practices.<\/p>\n\n\t<\/div>\n\n\t\t\t\t\t\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/li>\n\t\t\t\t\t<\/ul>\n\t<\/div>\n\t\t\t<\/div>\n\t\t<\/div>\n\t\t<\/div>\n\t<\/div>\n<\/section>\n","protected":false},"excerpt":{"rendered":"<p>What Are Visual Agents? If you\u2019ve ever wished your chatbot could look you in the eye, smile, and hold a natural conversation, you\u2019re in the right place. D\u2011ID\u2019s Visual Agents make that possible. No cameras, no crews, just a few clicks (and the right prompts) in the Studio. Visual agents are interactive AI avatars that&#8230;<\/p>\n","protected":false},"author":22,"featured_media":10484,"comment_status":"closed","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"_acf_changed":true,"content-type":"","_uag_custom_page_level_css":"","footnotes":""},"categories":[111,85],"tags":[27,68,40,140,141,142],"class_list":["post-10471","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-d-id-agents","category-generative-ai","tag-ai-technology","tag-generative-ai","tag-interactive-marketing","tag-conversationalai-2","tag-interactiveavatars","tag-visualagents-2"],"acf":[],"yoast_head":"<!-- This site is optimized with the Yoast SEO Premium plugin v27.4 (Yoast SEO v27.5) - https:\/\/yoast.com\/product\/yoast-seo-premium-wordpress\/ -->\n<title>How to Build a D\u2011ID Visual Agent: A Prompt\u2011by\u2011Prompt Guide | D-ID<\/title>\n<meta name=\"description\" content=\"Learn to create a D-ID Visual Agent with our step-by-step prompt guide, simplifying the process for all skill levels.\" \/>\n<meta name=\"robots\" content=\"index, follow, max-snippet:-1, max-image-preview:large, max-video-preview:-1\" \/>\n<link rel=\"canonical\" href=\"https:\/\/www.d-id.com\/blog\/how-to-build-a-did-visual-agent-a-promptbyprompt-guide\/\" \/>\n<meta property=\"og:locale\" content=\"en_US\" \/>\n<meta property=\"og:type\" content=\"article\" \/>\n<meta property=\"og:title\" content=\"How to Build a D\u2011ID Visual Agent: A Prompt\u2011by\u2011Prompt Guide\" \/>\n<meta property=\"og:description\" content=\"Learn to create a D-ID Visual Agent with our step-by-step prompt guide, simplifying the process for all skill levels.\" \/>\n<meta property=\"og:url\" content=\"https:\/\/www.d-id.com\/blog\/how-to-build-a-did-visual-agent-a-promptbyprompt-guide\/\" \/>\n<meta property=\"og:site_name\" content=\"D-ID\" \/>\n<meta property=\"article:publisher\" content=\"https:\/\/www.facebook.com\/deidentification\/\" \/>\n<meta property=\"article:published_time\" content=\"2025-07-31T11:12:50+00:00\" \/>\n<meta property=\"article:modified_time\" content=\"2026-02-18T12:59:27+00:00\" \/>\n<meta property=\"og:image\" content=\"https:\/\/www.d-id.com\/wp-content\/uploads\/2025\/07\/Cover-image-Visual-Agent-creation-guide.png\" \/>\n\t<meta property=\"og:image:width\" content=\"1200\" \/>\n\t<meta property=\"og:image:height\" content=\"628\" \/>\n\t<meta property=\"og:image:type\" content=\"image\/png\" \/>\n<meta name=\"author\" content=\"Ron Friedman\" \/>\n<meta name=\"twitter:card\" content=\"summary_large_image\" \/>\n<meta name=\"twitter:creator\" content=\"@D_ID_\" \/>\n<meta name=\"twitter:site\" content=\"@D_ID_\" \/>\n<meta name=\"twitter:label1\" content=\"Written by\" \/>\n\t<meta name=\"twitter:data1\" content=\"Ron Friedman\" \/>\n\t<meta name=\"twitter:label2\" content=\"Est. reading time\" \/>\n\t<meta name=\"twitter:data2\" content=\"14 minutes\" \/>\n<script type=\"application\/ld+json\" class=\"yoast-schema-graph\">{\"@context\":\"https:\\\/\\\/schema.org\",\"@graph\":[{\"@type\":\"Article\",\"@id\":\"https:\\\/\\\/www.d-id.com\\\/blog\\\/how-to-build-a-did-visual-agent-a-promptbyprompt-guide\\\/#article\",\"isPartOf\":{\"@id\":\"https:\\\/\\\/www.d-id.com\\\/blog\\\/how-to-build-a-did-visual-agent-a-promptbyprompt-guide\\\/\"},\"author\":{\"name\":\"Ron Friedman\",\"@id\":\"https:\\\/\\\/www.d-id.com\\\/#\\\/schema\\\/person\\\/1f786d04e5472fbc4fb24728bf5a1cb3\"},\"headline\":\"How to Build a D\u2011ID Visual Agent: A Prompt\u2011by\u2011Prompt Guide\",\"datePublished\":\"2025-07-31T11:12:50+00:00\",\"dateModified\":\"2026-02-18T12:59:27+00:00\",\"mainEntityOfPage\":{\"@id\":\"https:\\\/\\\/www.d-id.com\\\/blog\\\/how-to-build-a-did-visual-agent-a-promptbyprompt-guide\\\/\"},\"wordCount\":2916,\"publisher\":{\"@id\":\"https:\\\/\\\/www.d-id.com\\\/#organization\"},\"image\":{\"@id\":\"https:\\\/\\\/www.d-id.com\\\/blog\\\/how-to-build-a-did-visual-agent-a-promptbyprompt-guide\\\/#primaryimage\"},\"thumbnailUrl\":\"https:\\\/\\\/www.d-id.com\\\/wp-content\\\/uploads\\\/2025\\\/07\\\/Cover-image-Visual-Agent-creation-guide.png\",\"keywords\":[\"#AItechnology\",\"#GenerativeAi\",\"#interactivemarketing\",\"conversationalai\",\"interactiveavatars\",\"visualagents\"],\"articleSection\":[\"D-ID Agents\",\"Generative AI\"],\"inLanguage\":\"en-US\"},{\"@type\":\"WebPage\",\"@id\":\"https:\\\/\\\/www.d-id.com\\\/blog\\\/how-to-build-a-did-visual-agent-a-promptbyprompt-guide\\\/\",\"url\":\"https:\\\/\\\/www.d-id.com\\\/blog\\\/how-to-build-a-did-visual-agent-a-promptbyprompt-guide\\\/\",\"name\":\"How to Build a D\u2011ID Visual Agent: A Prompt\u2011by\u2011Prompt Guide | D-ID\",\"isPartOf\":{\"@id\":\"https:\\\/\\\/www.d-id.com\\\/#website\"},\"primaryImageOfPage\":{\"@id\":\"https:\\\/\\\/www.d-id.com\\\/blog\\\/how-to-build-a-did-visual-agent-a-promptbyprompt-guide\\\/#primaryimage\"},\"image\":{\"@id\":\"https:\\\/\\\/www.d-id.com\\\/blog\\\/how-to-build-a-did-visual-agent-a-promptbyprompt-guide\\\/#primaryimage\"},\"thumbnailUrl\":\"https:\\\/\\\/www.d-id.com\\\/wp-content\\\/uploads\\\/2025\\\/07\\\/Cover-image-Visual-Agent-creation-guide.png\",\"datePublished\":\"2025-07-31T11:12:50+00:00\",\"dateModified\":\"2026-02-18T12:59:27+00:00\",\"description\":\"Learn to create a D-ID Visual Agent with our step-by-step prompt guide, simplifying the process for all skill levels.\",\"breadcrumb\":{\"@id\":\"https:\\\/\\\/www.d-id.com\\\/blog\\\/how-to-build-a-did-visual-agent-a-promptbyprompt-guide\\\/#breadcrumb\"},\"inLanguage\":\"en-US\",\"potentialAction\":[{\"@type\":\"ReadAction\",\"target\":[\"https:\\\/\\\/www.d-id.com\\\/blog\\\/how-to-build-a-did-visual-agent-a-promptbyprompt-guide\\\/\"]}]},{\"@type\":\"ImageObject\",\"inLanguage\":\"en-US\",\"@id\":\"https:\\\/\\\/www.d-id.com\\\/blog\\\/how-to-build-a-did-visual-agent-a-promptbyprompt-guide\\\/#primaryimage\",\"url\":\"https:\\\/\\\/www.d-id.com\\\/wp-content\\\/uploads\\\/2025\\\/07\\\/Cover-image-Visual-Agent-creation-guide.png\",\"contentUrl\":\"https:\\\/\\\/www.d-id.com\\\/wp-content\\\/uploads\\\/2025\\\/07\\\/Cover-image-Visual-Agent-creation-guide.png\",\"width\":1200,\"height\":628,\"caption\":\"How to Build a D\u2011ID Visual Agent: A Prompt\u2011by\u2011Prompt Guide\"},{\"@type\":\"BreadcrumbList\",\"@id\":\"https:\\\/\\\/www.d-id.com\\\/blog\\\/how-to-build-a-did-visual-agent-a-promptbyprompt-guide\\\/#breadcrumb\",\"itemListElement\":[{\"@type\":\"ListItem\",\"position\":1,\"name\":\"Home\",\"item\":\"https:\\\/\\\/www.d-id.com\\\/\"},{\"@type\":\"ListItem\",\"position\":2,\"name\":\"How to Build a D\u2011ID Visual Agent: A Prompt\u2011by\u2011Prompt Guide\"}]},{\"@type\":\"WebSite\",\"@id\":\"https:\\\/\\\/www.d-id.com\\\/#website\",\"url\":\"https:\\\/\\\/www.d-id.com\\\/\",\"name\":\"D-ID\",\"description\":\"Create AI Videos, Interactive Avatars to engage your audience. Custom AI-powered digital people at scale for businesses and creators.\",\"publisher\":{\"@id\":\"https:\\\/\\\/www.d-id.com\\\/#organization\"},\"alternateName\":\"Interfaces, Evolved.\",\"potentialAction\":[{\"@type\":\"SearchAction\",\"target\":{\"@type\":\"EntryPoint\",\"urlTemplate\":\"https:\\\/\\\/www.d-id.com\\\/?s={search_term_string}\"},\"query-input\":{\"@type\":\"PropertyValueSpecification\",\"valueRequired\":true,\"valueName\":\"search_term_string\"}}],\"inLanguage\":\"en-US\"},{\"@type\":\"Organization\",\"@id\":\"https:\\\/\\\/www.d-id.com\\\/#organization\",\"name\":\"D-ID\",\"url\":\"https:\\\/\\\/www.d-id.com\\\/\",\"logo\":{\"@type\":\"ImageObject\",\"inLanguage\":\"en-US\",\"@id\":\"https:\\\/\\\/www.d-id.com\\\/#\\\/schema\\\/logo\\\/image\\\/\",\"url\":\"https:\\\/\\\/www.d-id.com\\\/wp-content\\\/uploads\\\/2023\\\/11\\\/d-id-logo-1.svg\",\"contentUrl\":\"https:\\\/\\\/www.d-id.com\\\/wp-content\\\/uploads\\\/2023\\\/11\\\/d-id-logo-1.svg\",\"width\":66,\"height\":53,\"caption\":\"D-ID\"},\"image\":{\"@id\":\"https:\\\/\\\/www.d-id.com\\\/#\\\/schema\\\/logo\\\/image\\\/\"},\"sameAs\":[\"https:\\\/\\\/www.facebook.com\\\/deidentification\\\/\",\"https:\\\/\\\/x.com\\\/D_ID_\"]},{\"@type\":\"Person\",\"@id\":\"https:\\\/\\\/www.d-id.com\\\/#\\\/schema\\\/person\\\/1f786d04e5472fbc4fb24728bf5a1cb3\",\"name\":\"Ron Friedman\",\"description\":\"Ron Friedman is a creative marketing leader and former journalist with a passion for storytelling at the intersection of technology, ethics, and imagination. As Head of Content &amp; Creative Marketing at D-ID, he leads strategy and execution across high-impact campaigns, thought leadership, and AI-powered content innovation. With a Master\u2019s in Journalism from the University of British Columbia and a career spanning media, mission-driven nonprofits, and cutting-edge startups, Ron brings both editorial rigor and startup agility. Off the clock, he\u2019s fueled by travel, whether it\u2019s to foreign cities irl, or to strange planets or alternate universes as a lifelong science fiction fan.\",\"jobTitle\":\"Head of Content & Creative Marketing\",\"worksFor\":\"D-ID\",\"url\":\"https:\\\/\\\/www.d-id.com\\\/author\\\/ron-friedman\\\/\"}]}<\/script>\n<!-- \/ Yoast SEO Premium plugin. -->","yoast_head_json":{"title":"How to Build a D\u2011ID Visual Agent: A Prompt\u2011by\u2011Prompt Guide | D-ID","description":"Learn to create a D-ID Visual Agent with our step-by-step prompt guide, simplifying the process for all skill levels.","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:\/\/www.d-id.com\/blog\/how-to-build-a-did-visual-agent-a-promptbyprompt-guide\/","og_locale":"en_US","og_type":"article","og_title":"How to Build a D\u2011ID Visual Agent: A Prompt\u2011by\u2011Prompt Guide","og_description":"Learn to create a D-ID Visual Agent with our step-by-step prompt guide, simplifying the process for all skill levels.","og_url":"https:\/\/www.d-id.com\/blog\/how-to-build-a-did-visual-agent-a-promptbyprompt-guide\/","og_site_name":"D-ID","article_publisher":"https:\/\/www.facebook.com\/deidentification\/","article_published_time":"2025-07-31T11:12:50+00:00","article_modified_time":"2026-02-18T12:59:27+00:00","og_image":[{"width":1200,"height":628,"url":"https:\/\/www.d-id.com\/wp-content\/uploads\/2025\/07\/Cover-image-Visual-Agent-creation-guide.png","type":"image\/png"}],"author":"Ron Friedman","twitter_card":"summary_large_image","twitter_creator":"@D_ID_","twitter_site":"@D_ID_","twitter_misc":{"Written by":"Ron Friedman","Est. reading time":"14 minutes"},"schema":{"@context":"https:\/\/schema.org","@graph":[{"@type":"Article","@id":"https:\/\/www.d-id.com\/blog\/how-to-build-a-did-visual-agent-a-promptbyprompt-guide\/#article","isPartOf":{"@id":"https:\/\/www.d-id.com\/blog\/how-to-build-a-did-visual-agent-a-promptbyprompt-guide\/"},"author":{"name":"Ron Friedman","@id":"https:\/\/www.d-id.com\/#\/schema\/person\/1f786d04e5472fbc4fb24728bf5a1cb3"},"headline":"How to Build a D\u2011ID Visual Agent: A Prompt\u2011by\u2011Prompt Guide","datePublished":"2025-07-31T11:12:50+00:00","dateModified":"2026-02-18T12:59:27+00:00","mainEntityOfPage":{"@id":"https:\/\/www.d-id.com\/blog\/how-to-build-a-did-visual-agent-a-promptbyprompt-guide\/"},"wordCount":2916,"publisher":{"@id":"https:\/\/www.d-id.com\/#organization"},"image":{"@id":"https:\/\/www.d-id.com\/blog\/how-to-build-a-did-visual-agent-a-promptbyprompt-guide\/#primaryimage"},"thumbnailUrl":"https:\/\/www.d-id.com\/wp-content\/uploads\/2025\/07\/Cover-image-Visual-Agent-creation-guide.png","keywords":["#AItechnology","#GenerativeAi","#interactivemarketing","conversationalai","interactiveavatars","visualagents"],"articleSection":["D-ID Agents","Generative AI"],"inLanguage":"en-US"},{"@type":"WebPage","@id":"https:\/\/www.d-id.com\/blog\/how-to-build-a-did-visual-agent-a-promptbyprompt-guide\/","url":"https:\/\/www.d-id.com\/blog\/how-to-build-a-did-visual-agent-a-promptbyprompt-guide\/","name":"How to Build a D\u2011ID Visual Agent: A Prompt\u2011by\u2011Prompt Guide | D-ID","isPartOf":{"@id":"https:\/\/www.d-id.com\/#website"},"primaryImageOfPage":{"@id":"https:\/\/www.d-id.com\/blog\/how-to-build-a-did-visual-agent-a-promptbyprompt-guide\/#primaryimage"},"image":{"@id":"https:\/\/www.d-id.com\/blog\/how-to-build-a-did-visual-agent-a-promptbyprompt-guide\/#primaryimage"},"thumbnailUrl":"https:\/\/www.d-id.com\/wp-content\/uploads\/2025\/07\/Cover-image-Visual-Agent-creation-guide.png","datePublished":"2025-07-31T11:12:50+00:00","dateModified":"2026-02-18T12:59:27+00:00","description":"Learn to create a D-ID Visual Agent with our step-by-step prompt guide, simplifying the process for all skill levels.","breadcrumb":{"@id":"https:\/\/www.d-id.com\/blog\/how-to-build-a-did-visual-agent-a-promptbyprompt-guide\/#breadcrumb"},"inLanguage":"en-US","potentialAction":[{"@type":"ReadAction","target":["https:\/\/www.d-id.com\/blog\/how-to-build-a-did-visual-agent-a-promptbyprompt-guide\/"]}]},{"@type":"ImageObject","inLanguage":"en-US","@id":"https:\/\/www.d-id.com\/blog\/how-to-build-a-did-visual-agent-a-promptbyprompt-guide\/#primaryimage","url":"https:\/\/www.d-id.com\/wp-content\/uploads\/2025\/07\/Cover-image-Visual-Agent-creation-guide.png","contentUrl":"https:\/\/www.d-id.com\/wp-content\/uploads\/2025\/07\/Cover-image-Visual-Agent-creation-guide.png","width":1200,"height":628,"caption":"How to Build a D\u2011ID Visual Agent: A Prompt\u2011by\u2011Prompt Guide"},{"@type":"BreadcrumbList","@id":"https:\/\/www.d-id.com\/blog\/how-to-build-a-did-visual-agent-a-promptbyprompt-guide\/#breadcrumb","itemListElement":[{"@type":"ListItem","position":1,"name":"Home","item":"https:\/\/www.d-id.com\/"},{"@type":"ListItem","position":2,"name":"How to Build a D\u2011ID Visual Agent: A Prompt\u2011by\u2011Prompt Guide"}]},{"@type":"WebSite","@id":"https:\/\/www.d-id.com\/#website","url":"https:\/\/www.d-id.com\/","name":"D-ID","description":"Create AI Videos, Interactive Avatars to engage your audience. Custom AI-powered digital people at scale for businesses and creators.","publisher":{"@id":"https:\/\/www.d-id.com\/#organization"},"alternateName":"Interfaces, Evolved.","potentialAction":[{"@type":"SearchAction","target":{"@type":"EntryPoint","urlTemplate":"https:\/\/www.d-id.com\/?s={search_term_string}"},"query-input":{"@type":"PropertyValueSpecification","valueRequired":true,"valueName":"search_term_string"}}],"inLanguage":"en-US"},{"@type":"Organization","@id":"https:\/\/www.d-id.com\/#organization","name":"D-ID","url":"https:\/\/www.d-id.com\/","logo":{"@type":"ImageObject","inLanguage":"en-US","@id":"https:\/\/www.d-id.com\/#\/schema\/logo\/image\/","url":"https:\/\/www.d-id.com\/wp-content\/uploads\/2023\/11\/d-id-logo-1.svg","contentUrl":"https:\/\/www.d-id.com\/wp-content\/uploads\/2023\/11\/d-id-logo-1.svg","width":66,"height":53,"caption":"D-ID"},"image":{"@id":"https:\/\/www.d-id.com\/#\/schema\/logo\/image\/"},"sameAs":["https:\/\/www.facebook.com\/deidentification\/","https:\/\/x.com\/D_ID_"]},{"@type":"Person","@id":"https:\/\/www.d-id.com\/#\/schema\/person\/1f786d04e5472fbc4fb24728bf5a1cb3","name":"Ron Friedman","description":"Ron Friedman is a creative marketing leader and former journalist with a passion for storytelling at the intersection of technology, ethics, and imagination. As Head of Content &amp; Creative Marketing at D-ID, he leads strategy and execution across high-impact campaigns, thought leadership, and AI-powered content innovation. With a Master\u2019s in Journalism from the University of British Columbia and a career spanning media, mission-driven nonprofits, and cutting-edge startups, Ron brings both editorial rigor and startup agility. Off the clock, he\u2019s fueled by travel, whether it\u2019s to foreign cities irl, or to strange planets or alternate universes as a lifelong science fiction fan.","jobTitle":"Head of Content & Creative Marketing","worksFor":"D-ID","url":"https:\/\/www.d-id.com\/author\/ron-friedman\/"}]}},"uagb_featured_image_src":{"full":["https:\/\/www.d-id.com\/wp-content\/uploads\/2025\/07\/Cover-image-Visual-Agent-creation-guide.png",1200,628,false],"thumbnail":["https:\/\/www.d-id.com\/wp-content\/uploads\/2025\/07\/Cover-image-Visual-Agent-creation-guide-150x150.png",150,150,true],"medium":["https:\/\/www.d-id.com\/wp-content\/uploads\/2025\/07\/Cover-image-Visual-Agent-creation-guide-300x157.png",300,157,true],"medium_large":["https:\/\/www.d-id.com\/wp-content\/uploads\/2025\/07\/Cover-image-Visual-Agent-creation-guide-768x402.png",768,402,true],"large":["https:\/\/www.d-id.com\/wp-content\/uploads\/2025\/07\/Cover-image-Visual-Agent-creation-guide-1024x536.png",1024,536,true],"1536x1536":["https:\/\/www.d-id.com\/wp-content\/uploads\/2025\/07\/Cover-image-Visual-Agent-creation-guide.png",1200,628,false],"2048x2048":["https:\/\/www.d-id.com\/wp-content\/uploads\/2025\/07\/Cover-image-Visual-Agent-creation-guide.png",1200,628,false]},"uagb_author_info":{"display_name":"Ron Friedman","author_link":"https:\/\/www.d-id.com\/author\/ron-friedman\/"},"uagb_comment_info":0,"uagb_excerpt":"What Are Visual Agents? If you\u2019ve ever wished your chatbot could look you in the eye, smile, and hold a natural conversation, you\u2019re in the right place. D\u2011ID\u2019s Visual Agents make that possible. No cameras, no crews, just a few clicks (and the right prompts) in the Studio. Visual agents are interactive AI avatars that...","_links":{"self":[{"href":"https:\/\/www.d-id.com\/wp-json\/wp\/v2\/posts\/10471","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.d-id.com\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/www.d-id.com\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/www.d-id.com\/wp-json\/wp\/v2\/users\/22"}],"replies":[{"embeddable":true,"href":"https:\/\/www.d-id.com\/wp-json\/wp\/v2\/comments?post=10471"}],"version-history":[{"count":0,"href":"https:\/\/www.d-id.com\/wp-json\/wp\/v2\/posts\/10471\/revisions"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/www.d-id.com\/wp-json\/wp\/v2\/media\/10484"}],"wp:attachment":[{"href":"https:\/\/www.d-id.com\/wp-json\/wp\/v2\/media?parent=10471"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.d-id.com\/wp-json\/wp\/v2\/categories?post=10471"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.d-id.com\/wp-json\/wp\/v2\/tags?post=10471"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}