{"data":[{"overviewHtml":"\u003cp\u003eMiniMax H3 Highspeed turns a product photo, design sketch, or scene image into a short clip with motion, camera direction, and native audio. The OpenAny reference-image template accepts one image and produces a 4–15 second 768P video. Use it to check whether a product\u0026#39;s shape holds up as the camera moves. \u003ca href=\"https://www.minimax.io/blog/minimax-h3\"\u003eMiniMax H3 family overview\u003c/a\u003e · \u003ca href=\"https://openany.si/api/v1/key-market/model-prices?modelid=minimax-h3-highspeed\"\u003eOpenAny route parameters\u003c/a\u003e\u003c/p\u003e\u003ch2\u003eExample output\u003c/h2\u003e\u003cp\u003e\u003ca href=\"https://examples.aig.cloudflare.com/klingai/kling-v4/inputs/sneaker-sketch.jpg\"\u003ePencil sketch of a sneaker\u003c/a\u003e\u003c/p\u003e\u003cp\u003e\u003ca href=\"https://examples.aig.cloudflare.com/minimax/h3-highspeed/sketch-to-product-reveal.mp4\"\u003eWatch the Highspeed sketch-to-product demo\u003c/a\u003e\u003c/p\u003e\u003ctable\u003e\u003ctbody\u003e\u003ctr\u003e\u003cth\u003e\u003cp\u003eInput\u003c/p\u003e\u003c/th\u003e\u003cth\u003e\u003cp\u003eAction\u003c/p\u003e\u003c/th\u003e\u003cth\u003e\u003cp\u003eSample settings\u003c/p\u003e\u003c/th\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd\u003e\u003cp\u003eA sneaker sketch and a brief for a white upper, orange heel, and translucent sole.\u003c/p\u003e\u003c/td\u003e\u003ctd\u003e\u003cp\u003eThe shoe rises from the page onto a white turntable, and the camera slowly circles it.\u003c/p\u003e\u003c/td\u003e\u003ctd\u003e\u003cp\u003e5 seconds, 768P, adaptive aspect ratio\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003c/tbody\u003e\u003c/table\u003e\u003cp\u003eCloudflare published this Highspeed sample and used the sketch as a first frame. The OpenAny request below uses the same sketch as a visual reference. Compare the shape, material, and movement of the shoe. \u003ca href=\"https://developers.cloudflare.com/ai/models/minimax/h3-highspeed/#examples\"\u003eSample and original input\u003c/a\u003e\u003c/p\u003e\u003ch2\u003eChoose your settings\u003c/h2\u003e\u003ctable\u003e\u003ctbody\u003e\u003ctr\u003e\u003cth\u003e\u003cp\u003eParameter\u003c/p\u003e\u003c/th\u003e\u003cth\u003e\u003cp\u003eOpenAny template\u003c/p\u003e\u003c/th\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd\u003e\u003cp\u003eReference image\u003c/p\u003e\u003c/td\u003e\u003ctd\u003e\u003cp\u003eOne public HTTPS image with the \u003ccode\u003ereference_image\u003c/code\u003e role.\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd\u003e\u003cp\u003e\u003ccode\u003eduration\u003c/code\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd\u003e\u003cp\u003e4–15 seconds. Start with five seconds for a single action.\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd\u003e\u003cp\u003e\u003ccode\u003eresolution\u003c/code\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd\u003e\u003cp\u003e\u003ccode\u003e768P\u003c/code\u003e\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd\u003e\u003cp\u003e\u003ccode\u003easpect_ratio\u003c/code\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd\u003e\u003cp\u003e\u003ccode\u003eadaptive\u003c/code\u003e, \u003ccode\u003e21:9\u003c/code\u003e, \u003ccode\u003e16:9\u003c/code\u003e, \u003ccode\u003e4:3\u003c/code\u003e, \u003ccode\u003e1:1\u003c/code\u003e, \u003ccode\u003e3:4\u003c/code\u003e, or \u003ccode\u003e9:16\u003c/code\u003e\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd\u003e\u003cp\u003e\u003ccode\u003egenerate_audio\u003c/code\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd\u003e\u003cp\u003eFixed at \u003ccode\u003etrue\u003c/code\u003e. Describe the sound you want in the prompt.\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd\u003e\u003cp\u003e\u003ccode\u003en\u003c/code\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd\u003e\u003cp\u003eFixed at 1.\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003c/tbody\u003e\u003c/table\u003e\u003cp\u003eStart with \u003ccode\u003eadaptive\u003c/code\u003e to see how the source composition carries over. For a set of vertical clips, choose \u003ccode\u003e9:16\u003c/code\u003e and ask the model to keep the product centered.\u003c/p\u003e\u003ch2\u003eQuickstart\u003c/h2\u003e\u003cp\u003eIn the OpenAny Playground, choose \u003cstrong\u003eReference image to video\u003c/strong\u003e. To use the API, send a POST request to \u003ccode\u003ehttps://api.openany.si/v1/videos\u003c/code\u003e with the header \u003ccode\u003eAuthorization: Bearer $OPENANY_API_KEY\u003c/code\u003e.\u003c/p\u003e\u003cpre\u003e\u003ccode class=\"language-json\"\u003e{\n  \u0026#34;model\u0026#34;: \u0026#34;minimax-h3-highspeed\u0026#34;,\n  \u0026#34;route\u0026#34;: { \u0026#34;provider\u0026#34;: \u0026#34;yir\u0026#34; },\n  \u0026#34;route_mode\u0026#34;: \u0026#34;balanced\u0026#34;,\n  \u0026#34;input\u0026#34;: {\n    \u0026#34;type\u0026#34;: \u0026#34;reference\u0026#34;,\n    \u0026#34;prompt\u0026#34;: \u0026#34;Use the sneaker sketch as the design reference. The pencil lines resolve into a white leather sneaker with an orange heel and a translucent sole. Reveal it on a slowly turning white studio platform. Keep the whole shoe in frame and add a subtle paper rustle.\u0026#34;,\n    \u0026#34;references\u0026#34;: [\n      {\n        \u0026#34;role\u0026#34;: \u0026#34;reference_image\u0026#34;,\n        \u0026#34;url\u0026#34;: \u0026#34;https://examples.aig.cloudflare.com/klingai/kling-v4/inputs/sneaker-sketch.jpg\u0026#34;\n      }\n    ]\n  },\n  \u0026#34;parameters\u0026#34;: {\n    \u0026#34;duration\u0026#34;: 5,\n    \u0026#34;resolution\u0026#34;: \u0026#34;768P\u0026#34;,\n    \u0026#34;aspect_ratio\u0026#34;: \u0026#34;adaptive\u0026#34;,\n    \u0026#34;generate_audio\u0026#34;: true,\n    \u0026#34;n\u0026#34;: 1\n  }\n}\n\u003c/code\u003e\u003c/pre\u003e\u003cp\u003ePoll \u003ccode\u003eGET /v1/videos/{task_id}\u003c/code\u003e using the returned task ID. Retrieve the video URL from the result\u0026#39;s file list when the task succeeds. Check whether the sole outline, laces, and orange heel stay consistent as the shoe turns.\u003c/p\u003e\u003ch2\u003eHighspeed or H3 Lite\u003c/h2\u003e\u003cp\u003eBoth OpenAny templates accept one reference image and can produce 768P video. Highspeed offers an adaptive aspect ratio. Lite adds a 480p option for checking motion and framing at a lower resolution. To compare the models, use the same image, prompt, duration, resolution, and aspect ratio.\u003c/p\u003e","officialPricing":{"unit":"second","prices":[{"match":{"kind":"tier","label":"768P · Reference image to video","inputMode":"r2v_voice","rangeStart":0,"resolution":"768p"},"price":"0.03","quantity":1}],"currency":"USD","quantity":1,"checkedAt":"2026-11-09","publisher":"MiniMax (promotional)"},"importedInterface":{"status":"available","protocol":"yir-standard","sourceUrl":"https://yir.ai/docs/models/minimax/minimax-h3-highspeed","upstreamModel":"minimax/minimax-h3-highspeed","formats":[{"exampleEditKey":"58327331d50eace2a993ebf51eb977f04186ade6811ba68d4703915fc39f3b8b","exampleVersion":"8b19d023409e20b42ed4560f79be8c09445bed2fb8f6b002b682cba7bf8fe581","jsonSchema":{"type":"object","required":["model","input","parameters"],"properties":{"input":{"type":"object","required":["type","prompt","references"],"properties":{"type":{"const":"reference"},"prompt":{"type":"string","minLength":1,"description":"Describe the scene, motion, and camera movement in English or your preferred language."},"references":{"type":"array","items":{"type":"object","required":["role","url"],"properties":{"url":{"type":"string","format":"uri","pattern":"^https://","maxLength":2048},"role":{"const":"reference_image"}},"additionalProperties":false},"maxItems":1,"minItems":1}},"additionalProperties":false},"model":{"type":"string","const":"minimax-h3-highspeed"},"route":{"type":"object","properties":{"provider":{"const":"yir"}}},"parameters":{"type":"object","properties":{"n":{"enum":[1],"type":"integer","default":1,"description":"Each submitted job generates one video."},"duration":{"type":"integer","default":5,"maximum":15,"minimum":4,"description":"Output length in whole seconds."},"resolution":{"enum":["768P"],"type":"string","default":"768P","description":"Output video resolution."},"aspect_ratio":{"enum":["adaptive","21:9","16:9","4:3","1:1","3:4","9:16"],"type":"string","default":"adaptive","description":"Aspect ratio of the output. Available values depend on the selected input mode."},"generate_audio":{"enum":[true],"type":"boolean","default":true,"description":"This model generates native audio. The value is fixed to true."}},"additionalProperties":false},"route_mode":{"type":"string","default":"balanced"}}},"id":"yir:minimax/minimax-h3-highspeed:reference","model":"minimax-h3-highspeed","provider":"yir","endpointPath":"/v1/videos","upstreamPath":"/v1/videos/generations","queryEndpointPath":"/v1/videos/{task_id}","upstreamQueryPath":"/v1/jobs/{task_id}","requestKind":"video","inputMode":"reference","displayName":"Reference image to video","description":"Submit a Yir Standard video job, then query its task ID to retrieve the result.","prompt":"Animate the reference image with a slow camera push and gentle natural movement. Keep the original composition recognizable.","systemPrompt":"","resolution":"768P","requestTemplate":{"input":{"prompt":"Animate the reference image with a slow camera push and gentle natural movement. Keep the original composition recognizable.","references":[{"role":"reference_image","url":"https://images.unsplash.com/photo-1473448912268-2022ce9509d8?w=800"}],"type":"reference"},"model":"minimax-h3-highspeed","parameters":{"aspect_ratio":"adaptive","duration":5,"generate_audio":true,"n":1,"resolution":"768P"},"route":{"provider":"yir"},"route_mode":"balanced"},"parameterSchema":[{"name":"prompt","label":"Prompt","description":"Describe the scene, movement, and camera behavior.","type":"textarea","valueType":"string","path":"/input/prompt","required":true,"defaultValue":"Animate the reference image with a slow camera push and gentle natural movement. Keep the original composition recognizable.","minLength":1},{"name":"references","label":"Reference image URL (JSON)","description":"Provide exactly one public HTTPS image URL with the reference_image role. Replace the sample image with your own if desired.","type":"json","path":"/input/references","required":true,"defaultValue":[{"role":"reference_image","url":"https://images.unsplash.com/photo-1473448912268-2022ce9509d8?w=800"}],"minItems":1,"maxItems":1,"referenceUrlOnly":true,"referenceRole":"reference_image","referenceConstraints":{"max_references":1,"min_references":1,"allowed_reference_roles":["reference_image"],"reference_counts_by_role":{"reference_image":{"maximum":1,"minimum":1}},"required_reference_roles":["reference_image"]}},{"name":"duration","label":"Duration (seconds)","description":"Output length in whole seconds.","type":"number","valueType":"integer","path":"/parameters/duration","required":false,"defaultValue":5,"min":4,"max":15},{"name":"resolution","label":"Resolution","description":"Output video resolution.","type":"select","valueType":"string","path":"/parameters/resolution","required":false,"defaultValue":"768P","options":["768P"],"readOnly":true,"allowedValues":["768P"]},{"name":"aspect_ratio","label":"Aspect ratio","description":"Aspect ratio of the output. Available values depend on the selected input mode.","type":"select","valueType":"string","path":"/parameters/aspect_ratio","required":false,"defaultValue":"adaptive","options":["adaptive","21:9","16:9","4:3","1:1","3:4","9:16"],"allowedValues":["adaptive","21:9","16:9","4:3","1:1","3:4","9:16"]},{"name":"generate_audio","label":"Generate audio","description":"This model generates native audio. The value is fixed to true.","type":"boolean","valueType":"boolean","path":"/parameters/generate_audio","required":false,"defaultValue":true,"readOnly":true,"allowedValues":[true]},{"name":"n","label":"Outputs per job","description":"Each submitted job generates one video.","type":"number","valueType":"integer","path":"/parameters/n","required":false,"defaultValue":1,"min":1,"max":1,"readOnly":true,"allowedValues":[1]}],"responseBody":{"id":"YOUR_TASK_ID","model":"minimax/minimax-h3-highspeed","object":"job","status":"queued","parameters":{"n":1,"duration":5,"resolution":"768P","aspect_ratio":"adaptive","generate_audio":true}},"queryResponseBody":{"id":"YOUR_TASK_ID","model":"minimax/minimax-h3-highspeed","object":"job","result":{"files":[{"url":"https://example.com/YOUR_GENERATED_VIDEO.mp4","media_type":"video/mp4"}],"availability":"available"},"status":"succeeded","parameters":{"n":1,"duration":5,"resolution":"768P","aspect_ratio":"adaptive","generate_audio":true}},"queryResponseOrigin":"derived_from_official","queryResponseContainsPlaceholders":true,"responseOrigin":"derived_from_official","responseContainsPlaceholders":true,"requestOrigin":"platform_template","sourceUrl":"https://yir.ai/docs/models/minimax/minimax-h3-highspeed","isDefault":true,"active":true,"sortOrder":0}],"missing":[]},"id":"b2e15ccf-8145-4fc8-a011-2eb5a1195961","routeId":"8e4b5bf5-34f8-43de-a70d-34153e1977de","model":"minimax-h3-highspeed","provider":"yir","modelSource":"platform","providerKind":"third_party","providerLogoUrl":"https://yir.ai/images/model-manufacturers/minimax.png","upstreamPath":[{"type":"generations","path":"/v1/videos/generations","method":"POST","default":true,"task_kind":"text2video"},{"type":"get","path":"/v1/jobs/{task_id}","method":"GET","default":true,"task_kind":"text2video"}],"curlExample":"","bodySchema":{"type":"object","oneOf":[{"type":"object","required":["model","input","parameters"],"properties":{"input":{"type":"object","required":["type","prompt","references"],"properties":{"type":{"const":"reference"},"prompt":{"type":"string","minLength":1,"description":"Describe the scene, motion, and camera movement in English or your preferred language."},"references":{"type":"array","items":{"type":"object","required":["role","url"],"properties":{"url":{"type":"string","format":"uri","pattern":"^https://","maxLength":2048},"role":{"const":"reference_image"}},"additionalProperties":false},"maxItems":1,"minItems":1}},"additionalProperties":false},"model":{"type":"string","const":"minimax-h3-highspeed"},"route":{"type":"object","properties":{"provider":{"const":"yir"}}},"parameters":{"type":"object","properties":{"n":{"enum":[1],"type":"integer","default":1,"description":"Each submitted job generates one video."},"duration":{"type":"integer","default":5,"maximum":15,"minimum":4,"description":"Output length in whole seconds."},"resolution":{"enum":["768P"],"type":"string","default":"768P","description":"Output video resolution."},"aspect_ratio":{"enum":["adaptive","21:9","16:9","4:3","1:1","3:4","9:16"],"type":"string","default":"adaptive","description":"Aspect ratio of the output. Available values depend on the selected input mode."},"generate_audio":{"enum":[true],"type":"boolean","default":true,"description":"This model generates native audio. The value is fixed to true."}},"additionalProperties":false},"route_mode":{"type":"string","default":"balanced"}}}]},"matchRules":{},"isLowestPrice":true,"isDefaultRoute":false,"routes":1,"prompt":30000,"completion":0,"imageCompletion":0,"partnerPriceMultiplierBps":10000,"partnerPrompt":30000,"partnerCompletion":0,"promptSource":30000,"completionSource":0,"imageCompletionSource":0,"currency":"USD","active":true,"sortOrder":0,"billingType":"duration","billingUnit":"second","modalities":["video"],"outModalities":["video"],"description":"Generate 768P video with native audio from a single reference image using the MiniMax H3 Highspeed API. Upload a product photo or design sketch and describe the subject’s movement, camera path, and ambient sound to create a 4–15 second clip. Try turntable shots, close-ups, and ad storyboards to see how the same product looks from different angles. Find video generation request examples, supported parameters, and usage pricing on the model page.\n```","descriptionTranslations":{"ar":"حوّل صورة مرجعية واحدة إلى فيديو بدقة 768P مع صوت مولّد باستخدام واجهة MiniMax H3 Highspeed API. ارفع صورة منتج أو رسماً تصميمياً، وصِف حركة العنصر ومسار الكاميرا وأصوات المكان لإنشاء مقطع من 4 إلى 15 ثانية. جرّب لقطات على منصة دوّارة ولقطات مقرّبة ومشاهد أولية لإعلان، وقارن شكل المنتج وملمسه وتكوين الصورة من زوايا مختلفة. تجد أمثلة طلبات توليد الفيديو والمعلمات المتاحة وأسعار الاستخدام في صفحة تفاصيل النموذج.","ja":"MiniMax H3 Highspeed API では、1 枚の参照画像から音声付きの 768P 動画を生成できます。商品写真やデザインスケッチをアップロードし、被写体の動き、カメラの軌道、環境音を指定して、4～15 秒のクリップを作成できます。同じ商品で回転台を使ったショットや接写、広告用の絵コンテを試し、角度ごとの輪郭や質感、構図を比較できます。モデル詳細ページで、動画生成のリクエスト例、利用可能なパラメータ、利用料金を確認できます。","zh":"通过 MiniMax H3 Highspeed API，用一张参考图生成 768P 有声视频。你可以上传商品照片或设计草图，用文字描述主体动作、镜头路径和环境声，生成 4–15 秒的短片。适合给同一件商品试做转台展示、细节特写和广告分镜，比较商品在不同镜头下的轮廓、材质和构图。接入前可以在详情页查看视频生成的请求示例、可用参数和用量价格。"},"supportedFeatures":[],"pricingTiers":[{"actualCompletion":0,"actualPrompt":30000,"catalogReferences":[],"completion":0,"importPricingMapping":{"status":"mapped","reviews":[],"sourcePricing":null},"inputMode":"r2v_voice","label":"768P · Reference image to video","partnerCompletion":0,"partnerPrompt":30000,"prompt":30000,"rangeStart":0,"resolution":"768p"}],"timePricingRules":[],"latency":218,"routeStatus":"active","operations":[],"operationRegistry":{"model":"minimax-h3-highspeed","operations":[],"availableModes":[]},"createdAt":"2026-11-09T03:58:18Z","updatedAt":"2026-11-09T06:26:20Z"}]}