{"id":1654,"date":"2026-03-25T09:14:33","date_gmt":"2026-03-25T09:14:33","guid":{"rendered":"https:\/\/wiro.ai\/blog\/?p=1654"},"modified":"2026-09-27T22:47:54","modified_gmt":"2026-09-27T22:47:54","slug":"seedance-v1-pro-fast-vs-wan-2-6-5-prompt-video-test","status":"publish","type":"post","link":"https:\/\/wiro.ai\/blog\/seedance-v1-pro-fast-vs-wan-2-6-5-prompt-video-test\/","title":{"rendered":"Seedance V1 Pro Fast vs Wan 2.6: 5 Prompt Video Test"},"content":{"rendered":"<h2>Seedance V1 Pro Fast vs Wan 2.6: the five-prompt test<\/h2>\n<p><strong>Seedance V1 Pro Fast vs Wan 2.6<\/strong> is a straight text-to-video comparison, not a claim that either model wins every category. Both were asked to make the same five-second, 720p, 16:9 clips from the same prompts. Audio and watermarks were off. One output was generated per prompt, so this is a practical look at these particular samples rather than a statistical benchmark.<\/p>\n<div class=\"wp-block-group\">\n<p><strong>Contents<\/strong><\/p>\n<ul>\n<li><a href=\"#setup\">Test setup<\/a><\/li>\n<li><a href=\"#results\">What the five outputs show<\/a><\/li>\n<li><a href=\"#speed\">Run time and cost<\/a><\/li>\n<li><a href=\"#choice\">Which model to pick<\/a><\/li>\n<\/ul>\n<\/div>\n<h2>Models tested<\/h2>\n<p>The comparison uses <a href=\"https:\/\/wiro.ai\/models\/bytedance\/seedance-v1-pro-fast\">Seedance V1 Pro Fast on Wiro<\/a> and <a href=\"https:\/\/wiro.ai\/models\/alibaba\/wan-2-6\">Wan 2.6 on Wiro<\/a>. Seedance accepts text or an input image, with 480p, 720p, or 1080p output, five- or ten-second duration, several aspect ratios, an optional seed, and a fixed-camera control. Wan 2.6 also handles text and image input. Its Wiro controls include standard or pro mode, 720P or 1080P, five-second duration, single or multi-shot mode, prompt expansion, an optional negative prompt, reference images, and optional audio for eligible image-to-video work.<\/p>\n<p>That matters when reading these clips. This test deliberately uses only prompt-to-video, a five-second duration, 720p, 16:9, a single shot, no audio, and no reference media. It does not test Wan&#8217;s reference workflow, multi-shot setting, or Seedance image-to-video mode.<\/p>\n<h2 id=\"setup\">Test setup<\/h2>\n<ul>\n<li>Five prompts, copied unchanged between models<\/li>\n<li>Text-to-video only; no source image or reference image<\/li>\n<li>Duration: 5 seconds<\/li>\n<li>Resolution: 720p<\/li>\n<li>Aspect ratio: 16:9<\/li>\n<li>Audio: off<\/li>\n<li>Watermark: off<\/li>\n<li>One generated output per prompt and model<\/li>\n<\/ul>\n<p>The prompt set was built to expose different failure modes. The night market asks for dense motion, rain, neon reflections, forward handheld camera movement, and a readable sign. The watch tests controlled product movement, reflections, and screen text. The turtle checks stable motion amid small moving particles. The tomato scene stresses hands, a blade, liquid, and shallow depth of field. The paper city asks for an intentional stop-motion cadence instead of accidental flicker.<\/p>\n<h2 id=\"results\">What the outputs show<\/h2>\n<h3>1. Rainy night market: text, crowd motion, and handheld movement<\/h3>\n<p><strong>Prompt:<\/strong> Handheld night market shot in Bangkok during heavy rain, neon reflections on wet pavement, steam from food stalls, sign with clear text MANGO STICKY RICE, camera walks forward through crowd, shallow depth of field, 35mm documentary look.<\/p>\n<table>\n<thead>\n<tr>\n<th>Seedance V1 Pro Fast<\/th>\n<th>Wan 2.6<\/th>\n<\/tr>\n<\/thead>\n<tbody>\n<tr>\n<td>\n<figure><video controls preload=\"metadata\" style=\"width:100%;height:auto;\"><source src=\"https:\/\/wiro.ai\/blog\/wp-content\/uploads\/2026\/03\/seedance-p1.mp4\" type=\"video\/mp4\" \/><\/video><figcaption>Seedance output for the rainy night-market prompt.<\/figcaption><\/figure>\n<\/td>\n<td>\n<figure><video controls preload=\"metadata\" style=\"width:100%;height:auto;\"><source src=\"https:\/\/wiro.ai\/blog\/wp-content\/uploads\/2026\/03\/wan-p1.mp4\" type=\"video\/mp4\" \/><\/video><figcaption>Wan output for the same rainy night-market prompt.<\/figcaption><\/figure>\n<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n<p>These samples are useful for judging the whole frame, not just the sign. Look at whether the forward move remains coherent as people cross the scene, whether rain and steam read as separate layers, and whether pavement reflections stay attached to the ground. The requested wording is a deliberately difficult constraint. A viewer should treat any legible text as a positive result, but not assume it will be exact or stable frame to frame.<\/p>\n<h3>2. Product orbit: camera control and materials<\/h3>\n<p><strong>Prompt:<\/strong> Studio tabletop product video of a silver smartwatch on a black acrylic stand, softbox lighting, slow orbit camera move, screen shows big text NOVA, clean gradient background, crisp reflections, realistic materials.<\/p>\n<table>\n<thead>\n<tr>\n<th>Seedance V1 Pro Fast<\/th>\n<th>Wan 2.6<\/th>\n<\/tr>\n<\/thead>\n<tbody>\n<tr>\n<td>\n<figure><video controls preload=\"metadata\" style=\"width:100%;height:auto;\"><source src=\"https:\/\/wiro.ai\/blog\/wp-content\/uploads\/2026\/03\/seedance-p2.mp4\" type=\"video\/mp4\" \/><\/video><figcaption>Seedance output for the product-orbit prompt.<\/figcaption><\/figure>\n<\/td>\n<td>\n<figure><video controls preload=\"metadata\" style=\"width:100%;height:auto;\"><source src=\"https:\/\/wiro.ai\/blog\/wp-content\/uploads\/2026\/03\/wan-p2.mp4\" type=\"video\/mp4\" \/><\/video><figcaption>Wan output for the same product-orbit prompt.<\/figcaption><\/figure>\n<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n<p>This pair is about restraint. The watch should remain one object while the camera orbits, the black stand should stay rigid, and highlights should travel plausibly across metal and acrylic. The NOVA display is another hard text request. For a real product ad, use this test as evidence that a clean single-object brief is sensible, then inspect every frame before approving a branded claim or product UI.<\/p>\n<h3>3. Underwater tracking: particle-heavy motion<\/h3>\n<p><strong>Prompt:<\/strong> Underwater wide shot of a sea turtle swimming through a coral reef, sun rays in water, bubbles and particles, slow tracking camera, natural color, realistic motion.<\/p>\n<table>\n<thead>\n<tr>\n<th>Seedance V1 Pro Fast<\/th>\n<th>Wan 2.6<\/th>\n<\/tr>\n<\/thead>\n<tbody>\n<tr>\n<td>\n<figure><video controls preload=\"metadata\" style=\"width:100%;height:auto;\"><source src=\"https:\/\/wiro.ai\/blog\/wp-content\/uploads\/2026\/03\/seedance-p3.mp4\" type=\"video\/mp4\" \/><\/video><figcaption>Seedance output for the underwater tracking prompt.<\/figcaption><\/figure>\n<\/td>\n<td>\n<figure><video controls preload=\"metadata\" style=\"width:100%;height:auto;\"><source src=\"https:\/\/wiro.ai\/blog\/wp-content\/uploads\/2026\/03\/wan-p3.mp4\" type=\"video\/mp4\" \/><\/video><figcaption>Wan output for the same underwater tracking prompt.<\/figcaption><\/figure>\n<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n<p>Water scenes reveal temporal problems quickly. The best pass keeps the turtle silhouette consistent, lets particles drift rather than pulse, and separates caustic light from noisy texture. This is less about a dramatic camera move than continuity under many small moving details.<\/p>\n<h3>4. Macro food prep: hands, edge detail, and fluid motion<\/h3>\n<p><strong>Prompt:<\/strong> Macro close-up of a chef slicing a ripe tomato on a wooden board, slow motion juice splash, sharp focus on knife edge, warm kitchen light, shallow depth of field.<\/p>\n<table>\n<thead>\n<tr>\n<th>Seedance V1 Pro Fast<\/th>\n<th>Wan 2.6<\/th>\n<\/tr>\n<\/thead>\n<tbody>\n<tr>\n<td>\n<figure><video controls preload=\"metadata\" style=\"width:100%;height:auto;\"><source src=\"https:\/\/wiro.ai\/blog\/wp-content\/uploads\/2026\/03\/seedance-p4.mp4\" type=\"video\/mp4\" \/><\/video><figcaption>Seedance output for the macro food-prep prompt.<\/figcaption><\/figure>\n<\/td>\n<td>\n<figure><video controls preload=\"metadata\" style=\"width:100%;height:auto;\"><source src=\"https:\/\/wiro.ai\/blog\/wp-content\/uploads\/2026\/03\/wan-p4.mp4\" type=\"video\/mp4\" \/><\/video><figcaption>Wan output for the same macro food-prep prompt.<\/figcaption><\/figure>\n<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n<p>Knife, hand, tomato, board, splash, and focus plane all compete here. Check whether the knife edge stays plausible through the cut and whether the tomato changes form naturally. Slow-motion wording cannot guarantee physically accurate fluid simulation, so this is a quality-control test for b-roll rather than evidence for a food demonstration.<\/p>\n<h3>5. Paper city: controlled stylization<\/h3>\n<p><strong>Prompt:<\/strong> Stop motion paper cutout animation, tiny cardboard city assembles itself on a desk, visible paper textures, overhead camera then gentle tilt, warm lamp light, playful motion.<\/p>\n<table>\n<thead>\n<tr>\n<th>Seedance V1 Pro Fast<\/th>\n<th>Wan 2.6<\/th>\n<\/tr>\n<\/thead>\n<tbody>\n<tr>\n<td>\n<figure><video controls preload=\"metadata\" style=\"width:100%;height:auto;\"><source src=\"https:\/\/wiro.ai\/blog\/wp-content\/uploads\/2026\/03\/seedance-p5.mp4\" type=\"video\/mp4\" \/><\/video><figcaption>Seedance output for the paper-city prompt.<\/figcaption><\/figure>\n<\/td>\n<td>\n<figure><video controls preload=\"metadata\" style=\"width:100%;height:auto;\"><source src=\"https:\/\/wiro.ai\/blog\/wp-content\/uploads\/2026\/03\/wan-p5.mp4\" type=\"video\/mp4\" \/><\/video><figcaption>Wan output for the same paper-city prompt.<\/figcaption><\/figure>\n<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n<p>This is the clearest style-control check. Good stop motion has deliberate stepwise movement while the paper texture, desk, and city geometry remain readable. Unwanted flicker looks different: it changes surfaces or objects without a visual reason. The gentle camera tilt also tests whether the model can add a move without losing the handmade look.<\/p>\n<h2 id=\"speed\">Run time and cost on Wiro<\/h2>\n<p>The historical task-detail responses for these ten outputs recorded provider elapsed time, not end-to-end wall-clock time. That excludes queueing and other platform overhead, so it should not be read as an SLA. Across these samples, Wan completed four of five runs in 37-41 seconds; Seedance ranged from 26-59 seconds.<\/p>\n<table>\n<thead>\n<tr>\n<th>Prompt<\/th>\n<th>Seedance elapsed<\/th>\n<th>Wan elapsed<\/th>\n<\/tr>\n<\/thead>\n<tbody>\n<tr>\n<td>Rainy night market<\/td>\n<td>52 s<\/td>\n<td>38 s<\/td>\n<\/tr>\n<tr>\n<td>Product orbit<\/td>\n<td>55 s<\/td>\n<td>41 s<\/td>\n<\/tr>\n<tr>\n<td>Underwater tracking<\/td>\n<td>49 s<\/td>\n<td>39 s<\/td>\n<\/tr>\n<tr>\n<td>Macro food prep<\/td>\n<td>26 s<\/td>\n<td>37 s<\/td>\n<\/tr>\n<tr>\n<td>Paper city<\/td>\n<td>59 s<\/td>\n<td>38 s<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n<p>Neither model&#8217;s current Wiro documentation exposes a per-output price for this exact 720p, five-second configuration, and the retained task details for these runs do not provide a cost field. No cost is estimated here. That is more useful than presenting a number that cannot be verified, especially because price can vary with model settings and future catalog changes.<\/p>\n<h2 id=\"choice\">When to pick each model<\/h2>\n<p>Pick <a href=\"https:\/\/wiro.ai\/models\/bytedance\/seedance-v1-pro-fast\">Seedance V1 Pro Fast<\/a> when the brief needs a text-only or image-led short clip and camera direction is central. Its Wiro options include 1080p, ten-second output, multiple ratios, a seed, and a fixed-camera switch. The official Seedance material also describes text-and-image generation and multi-shot storytelling, although those features were not tested here.<\/p>\n<p>Pick <a href=\"https:\/\/wiro.ai\/models\/alibaba\/wan-2-6\">Wan 2.6<\/a> when the production plan could benefit from its wider workflow controls: standard or pro mode, reference images, single or multi-shot generation, prompt expansion, a negative prompt, and optional audio on eligible image-to-video jobs. In this narrow test, its recorded provider times were more tightly clustered. That does not prove it will be faster for every mode or queue condition.<\/p>\n<p>For adjacent tests, see <a href=\"https:\/\/wiro.ai\/blog\/seedance-2-0-vs-seedance-v1-pro-fast-5-prompt-video-test\/\">Seedance 2.0 vs Seedance V1 Pro Fast<\/a>, <a href=\"https:\/\/wiro.ai\/blog\/8-prompts-for-vertical-video-with-wan-2-6\/\">8 Prompts for Vertical Video with Wan 2.6<\/a>, and <a href=\"https:\/\/wiro.ai\/blog\/ltx-video-vs-kling-vs-seedance\/\">LTX-Video vs Kling vs Seedance<\/a>.<\/p>\n<h2>Sources<\/h2>\n<ul>\n<li><a href=\"https:\/\/seed.bytedance.com\/en\/seedance\" target=\"_blank\" rel=\"noopener\">ByteDance Seedance official page<\/a><\/li>\n<li><a href=\"https:\/\/arxiv.org\/html\/2506.09113v1\" target=\"_blank\" rel=\"noopener\">Seedance 1.0 technical report on arXiv<\/a><\/li>\n<li><a href=\"https:\/\/github.com\/Wan-Video\" target=\"_blank\" rel=\"noopener\">Wan-Video on GitHub<\/a><\/li>\n<\/ul>\n<h2>Try the models<\/h2>\n<p>Run the same prompts on <a href=\"https:\/\/wiro.ai\/models\/bytedance\/seedance-v1-pro-fast\">Seedance V1 Pro Fast<\/a> and <a href=\"https:\/\/wiro.ai\/models\/alibaba\/wan-2-6\">Wan 2.6<\/a>, then compare the clips at full size before deciding which fits the shot.<\/p>\n","protected":false},"excerpt":{"rendered":"<p>Seedance V1 Pro Fast vs Wan 2.6: the five-prompt test Seedance V1 Pro Fast vs Wan 2.6 is a straight text-to-video comparison,&hellip;<\/p>\n","protected":false},"author":4,"featured_media":1653,"comment_status":"closed","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[51],"tags":[139,88,58,57],"class_list":["post-1654","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-model-comparison","tag-alibaba","tag-bytedance","tag-image-to-video","tag-text-to-video"],"_links":{"self":[{"href":"https:\/\/wiro.ai\/blog\/wp-json\/wp\/v2\/posts\/1654","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/wiro.ai\/blog\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/wiro.ai\/blog\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/wiro.ai\/blog\/wp-json\/wp\/v2\/users\/4"}],"replies":[{"embeddable":true,"href":"https:\/\/wiro.ai\/blog\/wp-json\/wp\/v2\/comments?post=1654"}],"version-history":[{"count":2,"href":"https:\/\/wiro.ai\/blog\/wp-json\/wp\/v2\/posts\/1654\/revisions"}],"predecessor-version":[{"id":4322,"href":"https:\/\/wiro.ai\/blog\/wp-json\/wp\/v2\/posts\/1654\/revisions\/4322"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/wiro.ai\/blog\/wp-json\/wp\/v2\/media\/1653"}],"wp:attachment":[{"href":"https:\/\/wiro.ai\/blog\/wp-json\/wp\/v2\/media?parent=1654"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/wiro.ai\/blog\/wp-json\/wp\/v2\/categories?post=1654"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/wiro.ai\/blog\/wp-json\/wp\/v2\/tags?post=1654"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}