[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"docs-content":3,"/ai-video-generator":1192,"/ai-image-generator":1192,"/pricing":1192,"/docs/ai-image-generator":1221},{"id":4,"title":5,"body":6,"category":1180,"date":1181,"description":1182,"extension":1183,"image":1184,"locale":1185,"meta":1186,"navigation":1191,"og":1192,"path":1193,"questions":1194,"seo":1213,"stem":1214,"tags":1215,"__hash__":1220},"content/en/docs/ai-video-generator.md","AI Video Generator: Complete Guide",{"type":7,"value":8,"toc":1128},"minimark",[9,19,24,27,32,256,260,267,273,279,285,291,295,298,302,323,327,354,358,361,365,390,394,420,424,427,431,542,546,610,614,617,621,659,662,666,669,673,676,680,706,710,713,717,720,731,735,738,742,745,749,763,766,770,774,777,788,791,795,798,818,822,825,829,858,862,865,869,872,876,906,910,954,958,978,982,985,1046,1050,1076,1084,1088,1094,1100,1106,1112,1118],[10,11,12,13,18],"p",{},"The ",[14,15,17],"a",{"href":16},"/ai-video-generator","AI Video Generator"," on Melies lets you create cinematic video clips using 8 AI models from the world's leading providers. From quick concept clips to production-quality footage with native audio, this guide covers every feature available.",[20,21,23],"h2",{"id":22},"available-models","Available Models",[10,25,26],{},"Melies offers 8 video generation models, each with distinct strengths in quality, duration, resolution, and special features.",[28,29,31],"h3",{"id":30},"model-comparison-table","Model Comparison Table",[33,34,35,66],"table",{},[36,37,38],"thead",{},[39,40,41,45,48,51,54,57,60,63],"tr",{},[42,43,44],"th",{},"Model",[42,46,47],{},"Provider",[42,49,50],{},"Speed",[42,52,53],{},"Quality",[42,55,56],{},"Credits",[42,58,59],{},"Resolution",[42,61,62],{},"Duration",[42,64,65],{},"Key Features",[67,68,69,96,122,143,164,187,210,232],"tbody",{},[39,70,71,75,78,81,84,87,90,93],{},[72,73,74],"td",{},"Veo 3.1",[72,76,77],{},"Google",[72,79,80],{},"Slow",[72,82,83],{},"Best",[72,85,86],{},"400",[72,88,89],{},"4K",[72,91,92],{},"4-8s",[72,94,95],{},"Native audio generation",[39,97,98,101,104,107,110,113,116,119],{},[72,99,100],{},"Kling v3 Pro",[72,102,103],{},"Kuaishou",[72,105,106],{},"Medium",[72,108,109],{},"High",[72,111,112],{},"100",[72,114,115],{},"HD",[72,117,118],{},"3-15s",[72,120,121],{},"Multi-shot, lip sync, Voice ID",[39,123,124,127,129,131,133,136,138,140],{},[72,125,126],{},"Kling O3 Standard",[72,128,103],{},[72,130,106],{},[72,132,109],{},[72,134,135],{},"80",[72,137,115],{},[72,139,118],{},[72,141,142],{},"Character elements",[39,144,145,148,150,152,155,158,160,162],{},[72,146,147],{},"Kling v3 Standard",[72,149,103],{},[72,151,106],{},[72,153,154],{},"Good",[72,156,157],{},"60",[72,159,115],{},[72,161,118],{},[72,163,142],{},[39,165,166,169,172,174,176,178,181,184],{},[72,167,168],{},"Seedance v1 Pro",[72,170,171],{},"ByteDance",[72,173,106],{},[72,175,154],{},[72,177,135],{},[72,179,180],{},"1080p",[72,182,183],{},"2-12s",[72,185,186],{},"-",[39,188,189,192,195,197,199,202,205,208],{},[72,190,191],{},"Hailuo 02",[72,193,194],{},"MiniMax",[72,196,106],{},[72,198,154],{},[72,200,201],{},"70",[72,203,204],{},"512P/768P",[72,206,207],{},"6s or 10s",[72,209,186],{},[39,211,212,215,218,220,222,224,227,230],{},[72,213,214],{},"WAN v2.2",[72,216,217],{},"Alibaba",[72,219,106],{},[72,221,154],{},[72,223,157],{},[72,225,226],{},"480-720p",[72,228,229],{},"~5s",[72,231,186],{},[39,233,234,237,240,243,245,248,250,253],{},[72,235,236],{},"LTX 2 Pro",[72,238,239],{},"Lightricks",[72,241,242],{},"Fast",[72,244,154],{},[72,246,247],{},"50",[72,249,115],{},[72,251,252],{},"~4.8s",[72,254,255],{},"Camera movement presets (LoRA)",[28,257,259],{"id":258},"choosing-the-right-model","Choosing the Right Model",[10,261,262,266],{},[263,264,265],"strong",{},"For the highest quality",": Veo 3.1 produces the best visual quality at 4K resolution and includes native audio generation. At 400 credits per clip, it is best reserved for hero shots and final production footage.",[10,268,269,272],{},[263,270,271],{},"For versatile filmmaking",": Kling v3 Pro is the most feature-rich model with multi-shot sequences, lip sync, and Voice ID support. It offers the longest duration (up to 15 seconds) at a reasonable 100 credits.",[10,274,275,278],{},[263,276,277],{},"For character-driven scenes",": Kling O3 Standard and Kling v3 Standard both support character elements for maintaining visual consistency at 80 and 60 credits respectively.",[10,280,281,284],{},[263,282,283],{},"For fast iteration",": LTX 2 Pro is the fastest model at just 50 credits, with built-in camera movement presets. Great for previewing scenes and exploring ideas.",[10,286,287,290],{},[263,288,289],{},"For budget-friendly clips",": WAN v2.2 (60 credits) and LTX 2 Pro (50 credits) are the most affordable options for generating video concepts.",[20,292,294],{"id":293},"text-to-video","Text-to-Video",[10,296,297],{},"Text-to-video is the simplest way to generate a clip. Write a description of your scene and the model creates a video from scratch.",[28,299,301],{"id":300},"how-to-generate-video-from-text","How to Generate Video from Text",[303,304,305,311,314,317,320],"ol",{},[306,307,308,309],"li",{},"Open the ",[14,310,17],{"href":16},[306,312,313],{},"Select your preferred model",[306,315,316],{},"Write a detailed scene description in the prompt field",[306,318,319],{},"Choose duration and resolution settings (options vary by model)",[306,321,322],{},"Click generate",[28,324,326],{"id":325},"tips-for-better-text-to-video-results","Tips for Better Text-to-Video Results",[328,329,330,336,342,348],"ul",{},[306,331,332,335],{},[263,333,334],{},"Be specific about movement",": Describe what is happening in the scene, not just what it looks like. \"A woman walks slowly through a rainy street, looking over her shoulder\" works better than \"woman in rainy street.\"",[306,337,338,341],{},[263,339,340],{},"Include camera direction",": Mention camera movements like \"slow dolly forward\" or \"camera pans left\" to guide the composition.",[306,343,344,347],{},[263,345,346],{},"Set the mood",": Include lighting and atmosphere details. \"Dimly lit with warm amber tones\" gives the model clear visual direction.",[306,349,350,353],{},[263,351,352],{},"Keep it focused",": Each clip should describe one clear action or moment. Avoid cramming multiple actions into a single prompt.",[20,355,357],{"id":356},"image-to-video","Image-to-Video",[10,359,360],{},"Image-to-video lets you animate a still image, turning it into a moving clip. This workflow gives you much more control over the visual result because you start with a known frame.",[28,362,364],{"id":363},"how-to-generate-video-from-an-image","How to Generate Video from an Image",[303,366,367,375,379,382,385,388],{},[306,368,369,370,374],{},"Generate an image using the ",[14,371,373],{"href":372},"/ai-image-generator","AI Image Generator"," or upload your own",[306,376,308,377],{},[14,378,17],{"href":16},[306,380,381],{},"Upload or select your starting image",[306,383,384],{},"Write a prompt describing what should happen in the video (the movement, action, camera motion)",[306,386,387],{},"Select your model and settings",[306,389,322],{},[28,391,393],{"id":392},"why-use-image-to-video","Why Use Image-to-Video",[328,395,396,402,408,414],{},[306,397,398,401],{},[263,399,400],{},"Character consistency",": Generate a character image first, then animate it - the video will closely match the starting frame",[306,403,404,407],{},[263,405,406],{},"Precise framing",": Get the exact composition you want in the first frame before adding motion",[306,409,410,413],{},[263,411,412],{},"Storyboard to film",": Turn storyboard frames into animated sequences",[306,415,416,419],{},[263,417,418],{},"Cost efficiency",": Iterate on cheap images first (2 credits with Flux Schnell), then spend video credits only on frames you are happy with",[20,421,423],{"id":422},"duration-and-resolution-options","Duration and Resolution Options",[10,425,426],{},"Each model has different duration and resolution capabilities.",[28,428,430],{"id":429},"duration-by-model","Duration by Model",[33,432,433,448],{},[36,434,435],{},[39,436,437,439,442,445],{},[42,438,44],{},[42,440,441],{},"Minimum",[42,443,444],{},"Maximum",[42,446,447],{},"Notes",[67,449,450,463,476,486,496,508,521,532],{},[39,451,452,454,457,460],{},[72,453,74],{},[72,455,456],{},"4s",[72,458,459],{},"8s",[72,461,462],{},"Includes native audio",[39,464,465,467,470,473],{},[72,466,100],{},[72,468,469],{},"3s",[72,471,472],{},"15s",[72,474,475],{},"Longest single-clip duration",[39,477,478,480,482,484],{},[72,479,126],{},[72,481,469],{},[72,483,472],{},[72,485,186],{},[39,487,488,490,492,494],{},[72,489,147],{},[72,491,469],{},[72,493,472],{},[72,495,186],{},[39,497,498,500,503,506],{},[72,499,168],{},[72,501,502],{},"2s",[72,504,505],{},"12s",[72,507,186],{},[39,509,510,512,515,518],{},[72,511,191],{},[72,513,514],{},"6s",[72,516,517],{},"10s",[72,519,520],{},"Fixed at 6s or 10s",[39,522,523,525,527,529],{},[72,524,214],{},[72,526,229],{},[72,528,229],{},[72,530,531],{},"Fixed duration",[39,533,534,536,538,540],{},[72,535,236],{},[72,537,252],{},[72,539,252],{},[72,541,531],{},[28,543,545],{"id":544},"resolution-by-model","Resolution by Model",[33,547,548,557],{},[36,549,550],{},[39,551,552,554],{},[42,553,44],{},[42,555,556],{},"Resolution Options",[67,558,559,566,572,578,584,590,597,604],{},[39,560,561,563],{},[72,562,74],{},[72,564,565],{},"Up to 4K",[39,567,568,570],{},[72,569,100],{},[72,571,115],{},[39,573,574,576],{},[72,575,126],{},[72,577,115],{},[39,579,580,582],{},[72,581,147],{},[72,583,115],{},[39,585,586,588],{},[72,587,168],{},[72,589,180],{},[39,591,592,594],{},[72,593,191],{},[72,595,596],{},"512P or 768P",[39,598,599,601],{},[72,600,214],{},[72,602,603],{},"480p to 720p",[39,605,606,608],{},[72,607,236],{},[72,609,115],{},[20,611,613],{"id":612},"camera-movement-presets","Camera Movement Presets",[10,615,616],{},"LTX 2 Pro includes built-in camera movement presets powered by LoRA (Low-Rank Adaptation) technology. Instead of describing camera motion in your prompt, you can select a preset for precise, consistent results.",[28,618,620],{"id":619},"available-presets","Available Presets",[33,622,623,633],{},[36,624,625],{},[39,626,627,630],{},[42,628,629],{},"Preset",[42,631,632],{},"Description",[67,634,635,643,651],{},[39,636,637,640],{},[72,638,639],{},"Dolly",[72,641,642],{},"Camera moves forward or backward along a track, creating depth",[39,644,645,648],{},[72,646,647],{},"Jib",[72,649,650],{},"Camera moves vertically up or down, simulating a crane shot",[39,652,653,656],{},[72,654,655],{},"Static",[72,657,658],{},"Camera remains fixed, focusing entirely on subject movement",[10,660,661],{},"These presets are exclusive to LTX 2 Pro and are applied automatically during generation. They produce smoother, more predictable camera motion than prompt-based directions.",[20,663,665],{"id":664},"multi-shot-sequences","Multi-Shot Sequences",[10,667,668],{},"Kling v3 Pro supports multi-shot sequence generation, a powerful feature for creating connected scenes.",[28,670,672],{"id":671},"what-are-multi-shot-sequences","What Are Multi-Shot Sequences",[10,674,675],{},"Instead of generating individual clips and editing them together, multi-shot mode lets you generate several connected shots in a single generation. The model maintains visual consistency between shots, making transitions feel natural.",[28,677,679],{"id":678},"when-to-use-multi-shot","When to Use Multi-Shot",[328,681,682,688,694,700],{},[306,683,684,687],{},[263,685,686],{},"Dialogue scenes",": Generate shot and reverse-shot sequences",[306,689,690,693],{},[263,691,692],{},"Action sequences",": Create connected beats in an action scene",[306,695,696,699],{},[263,697,698],{},"Scene transitions",": Move from one framing to another smoothly",[306,701,702,705],{},[263,703,704],{},"Short narratives",": Tell a brief visual story in one generation",[20,707,709],{"id":708},"lip-sync-and-voice-id","Lip Sync and Voice ID",[10,711,712],{},"Kling v3 Pro includes lip sync and Voice ID capabilities for creating dialogue-driven scenes.",[28,714,716],{"id":715},"lip-sync","Lip Sync",[10,718,719],{},"Provide dialogue text and the model generates a character speaking those words with synchronized lip movements. This is valuable for:",[328,721,722,725,728],{},[306,723,724],{},"Creating talking-head scenes without live actors",[306,726,727],{},"Animating AI-generated characters with speech",[306,729,730],{},"Prototyping dialogue scenes before production",[28,732,734],{"id":733},"voice-id","Voice ID",[10,736,737],{},"Voice ID lets you maintain a consistent voice across multiple generated clips. Once you establish a voice for a character, you can reuse it in new generations, keeping your characters sounding the same throughout your project.",[20,739,741],{"id":740},"extending-videos","Extending Videos",[10,743,744],{},"Melies lets you extend generated videos beyond their initial duration. After generating a clip, you can extend it to continue the action or scene.",[28,746,748],{"id":747},"how-video-extension-works","How Video Extension Works",[303,750,751,754,757,760],{},[306,752,753],{},"Generate a video clip using any model",[306,755,756],{},"Select the extend option on the generated clip",[306,758,759],{},"The model generates additional frames that continue seamlessly from where the original clip ended",[306,761,762],{},"Repeat to build longer sequences",[10,764,765],{},"This is useful for creating clips longer than a single model's maximum duration, or for extending a particularly good generation that you want to continue.",[20,767,769],{"id":768},"audio-generation","Audio Generation",[28,771,773],{"id":772},"native-audio-with-veo-31","Native Audio with Veo 3.1",[10,775,776],{},"Veo 3.1 is the only model that generates synchronized audio alongside video. This includes:",[328,778,779,782,785],{},[306,780,781],{},"Ambient sounds matching the scene (rain, wind, crowds)",[306,783,784],{},"Sound effects tied to on-screen actions",[306,786,787],{},"Environmental audio that matches the visual setting",[10,789,790],{},"This makes Veo 3.1 uniquely suited for producing complete audiovisual clips without a separate audio step.",[28,792,794],{"id":793},"adding-audio-during-export","Adding Audio During Export",[10,796,797],{},"For all other models, you can add audio during the export process.",[328,799,800,806,812],{},[306,801,802,805],{},[263,803,804],{},"Music",": Add background music tracks with volume control",[306,807,808,811],{},[263,809,810],{},"Voice",": Layer voiceover or dialogue",[306,813,814,817],{},[263,815,816],{},"SFX",": Add sound effects",[20,819,821],{"id":820},"character-consistency-with-reference-images","Character Consistency with Reference Images",[10,823,824],{},"Maintaining consistent characters across multiple video clips is one of the biggest challenges in AI filmmaking. Melies offers several tools to help.",[28,826,828],{"id":827},"workflow-for-consistent-characters","Workflow for Consistent Characters",[303,830,831,840,846,852],{},[306,832,833,836,837,839],{},[263,834,835],{},"Create your character"," using the ",[14,838,373],{"href":372}," with a detailed description",[306,841,842,845],{},[263,843,844],{},"Extract the subject"," to isolate the character from the background",[306,847,848,851],{},[263,849,850],{},"Use as reference"," when generating new images and videos",[306,853,854,857],{},[263,855,856],{},"Select character element models"," like Kling O3 Standard or Kling v3 Standard for built-in character consistency",[28,859,861],{"id":860},"character-elements","Character Elements",[10,863,864],{},"Kling O3 Standard and Kling v3 Standard support character elements, which allow you to define persistent character traits that the model maintains across generations. This is more reliable than prompt-based consistency alone.",[20,866,868],{"id":867},"export-formats-and-settings","Export Formats and Settings",[10,870,871],{},"When your clips are ready, Melies offers flexible export options.",[28,873,875],{"id":874},"video-formats","Video Formats",[33,877,878,888],{},[36,879,880],{},[39,881,882,885],{},[42,883,884],{},"Format",[42,886,887],{},"Best For",[67,889,890,898],{},[39,891,892,895],{},[72,893,894],{},"MP4",[72,896,897],{},"Universal compatibility, social media, most players",[39,899,900,903],{},[72,901,902],{},"WebM",[72,904,905],{},"Web playback, smaller file sizes",[28,907,909],{"id":908},"export-settings","Export Settings",[33,911,912,922],{},[36,913,914],{},[39,915,916,919],{},[42,917,918],{},"Setting",[42,920,921],{},"Options",[67,923,924,931,938,946],{},[39,925,926,928],{},[72,927,59],{},[72,929,930],{},"480p, 720p, 1080p",[39,932,933,935],{},[72,934,53],{},[72,936,937],{},"Low (CRF 28), Medium (CRF 23), High (CRF 18)",[39,939,940,943],{},[72,941,942],{},"Frame Rate",[72,944,945],{},"1-60 FPS (default 25)",[39,947,948,951],{},[72,949,950],{},"Aspect Ratio",[72,952,953],{},"16:9, 3:2, 1:1, 2:3, 9:16",[28,955,957],{"id":956},"additional-export-options","Additional Export Options",[328,959,960,966,972],{},[306,961,962,965],{},[263,963,964],{},"Zoom mode",": Apply zoom effects to your final video",[306,967,968,971],{},[263,969,970],{},"Poster frame",": Set a specific frame as the video thumbnail",[306,973,974,977],{},[263,975,976],{},"Audio mixing",": Combine music, voice, and sound effects with individual volume controls",[20,979,981],{"id":980},"credit-costs-breakdown","Credit Costs Breakdown",[10,983,984],{},"Video generation costs more credits than images because of the computational complexity involved. Here is a quick reference.",[33,986,987,996],{},[36,988,989],{},[39,990,991,993],{},[42,992,44],{},[42,994,995],{},"Credits per Clip",[67,997,998,1004,1010,1016,1022,1028,1034,1040],{},[39,999,1000,1002],{},[72,1001,236],{},[72,1003,247],{},[39,1005,1006,1008],{},[72,1007,214],{},[72,1009,157],{},[39,1011,1012,1014],{},[72,1013,147],{},[72,1015,157],{},[39,1017,1018,1020],{},[72,1019,191],{},[72,1021,201],{},[39,1023,1024,1026],{},[72,1025,126],{},[72,1027,135],{},[39,1029,1030,1032],{},[72,1031,168],{},[72,1033,135],{},[39,1035,1036,1038],{},[72,1037,100],{},[72,1039,112],{},[39,1041,1042,1044],{},[72,1043,74],{},[72,1045,86],{},[28,1047,1049],{"id":1048},"cost-effective-video-workflow","Cost-Effective Video Workflow",[303,1051,1052,1058,1064,1070],{},[306,1053,1054,1057],{},[263,1055,1056],{},"Concept phase",": Use LTX 2 Pro (50 credits) or WAN v2.2 (60 credits) to test ideas quickly",[306,1059,1060,1063],{},[263,1061,1062],{},"Refinement",": Switch to Kling v3 Standard (60 credits) or Seedance v1 Pro (80 credits) for better quality",[306,1065,1066,1069],{},[263,1067,1068],{},"Production",": Use Kling v3 Pro (100 credits) for final clips that need features like lip sync or multi-shot",[306,1071,1072,1075],{},[263,1073,1074],{},"Hero shots",": Reserve Veo 3.1 (400 credits) for your most important scenes where 4K quality and native audio matter",[10,1077,1078,1079,1083],{},"Visit the ",[14,1080,1082],{"href":1081},"/pricing","Pricing"," page for current credit packages.",[20,1085,1087],{"id":1086},"tips-for-great-video-results","Tips for Great Video Results",[10,1089,1090,1093],{},[263,1091,1092],{},"Start with image-to-video",": Generating an image first and then animating it gives you far more control than text-to-video alone. Iterate on cheap images before spending video credits.",[10,1095,1096,1099],{},[263,1097,1098],{},"One action per clip",": Keep your prompts focused on a single movement or action. Multi-action prompts often produce confused results.",[10,1101,1102,1105],{},[263,1103,1104],{},"Use camera presets when available",": LTX 2 Pro's Dolly, Jib, and Static presets produce smoother camera motion than prompt-based direction.",[10,1107,1108,1111],{},[263,1109,1110],{},"Plan your shots like a real film",": Think in terms of establishing shots, close-ups, and reaction shots. Generate each as a separate clip and combine them in the export.",[10,1113,1114,1117],{},[263,1115,1116],{},"Match models to your needs",": Do not use Veo 3.1 for every clip. Use affordable models for B-roll and transition shots, and save premium models for key moments.",[10,1119,1120,1121,1123,1124,1127],{},"Ready to create your first video? Open the ",[14,1122,17],{"href":16}," and start generating. For still frames and character design, check out the ",[14,1125,373],{"href":1126},"/docs/ai-image-generator"," guide.",{"title":1129,"searchDepth":1130,"depth":1130,"links":1131},"",2,[1132,1137,1141,1145,1149,1152,1156,1160,1163,1167,1171,1176,1179],{"id":22,"depth":1130,"text":23,"children":1133},[1134,1136],{"id":30,"depth":1135,"text":31},3,{"id":258,"depth":1135,"text":259},{"id":293,"depth":1130,"text":294,"children":1138},[1139,1140],{"id":300,"depth":1135,"text":301},{"id":325,"depth":1135,"text":326},{"id":356,"depth":1130,"text":357,"children":1142},[1143,1144],{"id":363,"depth":1135,"text":364},{"id":392,"depth":1135,"text":393},{"id":422,"depth":1130,"text":423,"children":1146},[1147,1148],{"id":429,"depth":1135,"text":430},{"id":544,"depth":1135,"text":545},{"id":612,"depth":1130,"text":613,"children":1150},[1151],{"id":619,"depth":1135,"text":620},{"id":664,"depth":1130,"text":665,"children":1153},[1154,1155],{"id":671,"depth":1135,"text":672},{"id":678,"depth":1135,"text":679},{"id":708,"depth":1130,"text":709,"children":1157},[1158,1159],{"id":715,"depth":1135,"text":716},{"id":733,"depth":1135,"text":734},{"id":740,"depth":1130,"text":741,"children":1161},[1162],{"id":747,"depth":1135,"text":748},{"id":768,"depth":1130,"text":769,"children":1164},[1165,1166],{"id":772,"depth":1135,"text":773},{"id":793,"depth":1135,"text":794},{"id":820,"depth":1130,"text":821,"children":1168},[1169,1170],{"id":827,"depth":1135,"text":828},{"id":860,"depth":1135,"text":861},{"id":867,"depth":1130,"text":868,"children":1172},[1173,1174,1175],{"id":874,"depth":1135,"text":875},{"id":908,"depth":1135,"text":909},{"id":956,"depth":1135,"text":957},{"id":980,"depth":1130,"text":981,"children":1177},[1178],{"id":1048,"depth":1135,"text":1049},{"id":1086,"depth":1130,"text":1087},"docs","2026-03-03","Master the Melies AI Video Generator with 8 models including Veo 3.1, Kling v3, and more. Text-to-video, image-to-video, lip sync, and multi-shot guides.","md","/img/blog/video-model-selection.webp","en",{"sitemap":1187},{"loc":1188,"lastmod":1181,"changefreq":1189,"priority":1190},"/docs/ai-video-generator","monthly",0.8,true,null,"/en/docs/ai-video-generator",[1195,1198,1201,1204,1207,1210],{"question":1196,"answer":1197},"What AI video models does Melies support?","Melies supports 8 AI video models: Veo 3.1 (Google), Kling v3 Pro, Kling O3 Standard, Kling v3 Standard (Kuaishou), Seedance v1 Pro (ByteDance), Hailuo 02 (MiniMax), WAN v2.2 (Alibaba), and LTX 2 Pro (Lightricks).",{"question":1199,"answer":1200},"Can I generate video from a text prompt only?","Yes. All Melies video models support text-to-video generation. Write a description of your scene and the AI will generate a video clip. For better results, you can also provide a starting image using the image-to-video workflow.",{"question":1202,"answer":1203},"Which AI video model generates audio automatically?","Veo 3.1 by Google is the only model that generates native audio along with the video. Other models produce silent video that you can add sound to during export.",{"question":1205,"answer":1206},"How long can AI-generated videos be?","Video duration varies by model. Kling models support up to 15 seconds per clip. Seedance v1 Pro supports up to 12 seconds. Veo 3.1 generates 4-8 second clips. You can also extend videos to create longer sequences.",{"question":1208,"answer":1209},"What is multi-shot sequence generation?","Kling v3 Pro supports multi-shot sequences, allowing you to generate multiple connected shots in a single generation. This is useful for creating coherent scene transitions and short sequences without generating each shot separately.",{"question":1211,"answer":1212},"How does lip sync work in Melies?","Kling v3 Pro supports lip sync and Voice ID features. You can provide dialogue and the model will generate a character speaking those words with synchronized lip movements. Voice ID lets you maintain a consistent voice across scenes.",{"title":5,"description":1182},"en/docs/ai-video-generator",[1216,1217,1218,1219],"ai video generator","guide","documentation","video generation","mKntf7_iRIj3l6C7nha5l_yXfiMCB0xkFuswsmjCLug",{"id":1222,"title":1223,"body":1224,"category":1180,"date":1181,"description":2334,"extension":1183,"image":2335,"locale":1185,"meta":2336,"navigation":1191,"og":1192,"path":2338,"questions":2339,"seo":2358,"stem":2359,"tags":2360,"__hash__":2363},"content/en/docs/ai-image-generator.md","AI Image Generator: Complete Guide",{"type":7,"value":1225,"toc":2302},[1226,1231,1233,1236,1238,1529,1531,1540,1548,1557,1563,1578,1582,1585,1699,1702,1706,1709,1743,1746,1750,1753,1757,1775,1778,1804,1808,1811,1825,1828,1832,1835,1839,1842,1871,1875,1878,1907,1911,1914,1920,1926,1932,1936,1939,1971,1975,1978,1981,1985,1988,1991,1995,1998,2030,2034,2037,2040,2044,2047,2050,2054,2057,2060,2063,2066,2070,2073,2104,2107,2111,2114,2118,2132,2136,2156,2159,2161,2164,2248,2251,2256,2260,2266,2275,2281,2287,2293],[10,1227,12,1228,1230],{},[14,1229,373],{"href":372}," on Melies gives you access to 16 state-of-the-art models from leading AI providers. Whether you are creating concept art for a film, designing characters, or building storyboards, this guide covers everything you need to know.",[20,1232,23],{"id":22},[10,1234,1235],{},"Melies includes 16 image generation models. Each model has different strengths in speed, quality, and creative style. Here is a full comparison.",[28,1237,31],{"id":30},[33,1239,1240,1257],{},[36,1241,1242],{},[39,1243,1244,1246,1248,1250,1252,1254],{},[42,1245,44],{},[42,1247,47],{},[42,1249,50],{},[42,1251,53],{},[42,1253,56],{},[42,1255,1256],{},"Max Resolution",[67,1258,1259,1276,1293,1310,1326,1341,1358,1376,1392,1412,1429,1446,1466,1482,1497,1512],{},[39,1260,1261,1264,1266,1268,1270,1273],{},[72,1262,1263],{},"Nano Banana 2",[72,1265,77],{},[72,1267,242],{},[72,1269,83],{},[72,1271,1272],{},"15 (8-40 by res)",[72,1274,1275],{},"0.5K-4K",[39,1277,1278,1281,1283,1285,1287,1290],{},[72,1279,1280],{},"Nano Banana Pro",[72,1282,77],{},[72,1284,106],{},[72,1286,109],{},[72,1288,1289],{},"15",[72,1291,1292],{},"1K-4K",[39,1294,1295,1298,1300,1302,1304,1307],{},[72,1296,1297],{},"Nano Banana",[72,1299,77],{},[72,1301,242],{},[72,1303,154],{},[72,1305,1306],{},"10",[72,1308,1309],{},"1K",[39,1311,1312,1315,1317,1319,1321,1323],{},[72,1313,1314],{},"Seedream 4 Edit",[72,1316,171],{},[72,1318,106],{},[72,1320,109],{},[72,1322,1306],{},[72,1324,1325],{},"Auto/2K/4K",[39,1327,1328,1331,1333,1335,1337,1339],{},[72,1329,1330],{},"Seedream 4.5",[72,1332,171],{},[72,1334,106],{},[72,1336,109],{},[72,1338,1306],{},[72,1340,1325],{},[39,1342,1343,1346,1349,1351,1353,1355],{},[72,1344,1345],{},"Grok Imagine",[72,1347,1348],{},"xAI",[72,1350,106],{},[72,1352,109],{},[72,1354,1289],{},[72,1356,1357],{},"13 aspect ratios",[39,1359,1360,1363,1366,1368,1370,1373],{},[72,1361,1362],{},"Flux Kontext Max",[72,1364,1365],{},"Black Forest Labs",[72,1367,80],{},[72,1369,83],{},[72,1371,1372],{},"25",[72,1374,1375],{},"9 aspect ratios",[39,1377,1378,1381,1383,1385,1387,1390],{},[72,1379,1380],{},"Flux Kontext",[72,1382,1365],{},[72,1384,106],{},[72,1386,109],{},[72,1388,1389],{},"20",[72,1391,1375],{},[39,1393,1394,1401,1403,1405,1407,1409],{},[72,1395,1396,1397],{},"FLUX.2 ",[1398,1399,1400],"span",{},"max",[72,1402,1365],{},[72,1404,80],{},[72,1406,83],{},[72,1408,1372],{},[72,1410,1411],{},"Custom up to 14142px",[39,1413,1414,1419,1421,1423,1425,1427],{},[72,1415,1396,1416],{},[1398,1417,1418],{},"pro",[72,1420,1365],{},[72,1422,106],{},[72,1424,109],{},[72,1426,1389],{},[72,1428,1411],{},[39,1430,1431,1436,1438,1440,1442,1444],{},[72,1432,1396,1433],{},[1398,1434,1435],{},"flex",[72,1437,1365],{},[72,1439,242],{},[72,1441,154],{},[72,1443,1289],{},[72,1445,1411],{},[39,1447,1448,1453,1455,1458,1460,1463],{},[72,1449,1396,1450],{},[1398,1451,1452],{},"klein",[72,1454,1365],{},[72,1456,1457],{},"Fastest",[72,1459,154],{},[72,1461,1462],{},"5",[72,1464,1465],{},"Standard ratios",[39,1467,1468,1471,1473,1475,1477,1479],{},[72,1469,1470],{},"Flux Pro Ultra",[72,1472,1365],{},[72,1474,80],{},[72,1476,109],{},[72,1478,1389],{},[72,1480,1481],{},"Up to 2K",[39,1483,1484,1487,1489,1491,1493,1495],{},[72,1485,1486],{},"Flux Pro",[72,1488,1365],{},[72,1490,106],{},[72,1492,109],{},[72,1494,1289],{},[72,1496,1465],{},[39,1498,1499,1502,1504,1506,1508,1510],{},[72,1500,1501],{},"Flux Dev",[72,1503,1365],{},[72,1505,106],{},[72,1507,154],{},[72,1509,1306],{},[72,1511,1465],{},[39,1513,1514,1517,1519,1521,1524,1527],{},[72,1515,1516],{},"Flux Schnell",[72,1518,1365],{},[72,1520,1457],{},[72,1522,1523],{},"Basic",[72,1525,1526],{},"2",[72,1528,1465],{},[28,1530,259],{"id":258},[10,1532,1533,1536,1537,1539],{},[263,1534,1535],{},"For the best quality",": Nano Banana 2, Flux Kontext Max, and FLUX.2 ",[1398,1538,1400],{}," produce the most detailed and visually stunning images. Use these for hero shots and final production images.",[10,1541,1542,1544,1545,1547],{},[263,1543,283],{},": Flux Schnell (2 credits) and FLUX.2 ",[1398,1546,1452],{}," (5 credits) are the fastest and cheapest options. Perfect for brainstorming, exploring ideas, and quick concept drafts.",[10,1549,1550,1553,1554,1556],{},[263,1551,1552],{},"For balanced performance",": Seedream 4.5, Flux Kontext, and FLUX.2 ",[1398,1555,1418],{}," offer a strong balance between quality, speed, and credit cost.",[10,1558,1559,1562],{},[263,1560,1561],{},"For editing existing images",": Seedream 4 Edit is specifically designed for image editing workflows where you want to modify parts of an existing image.",[10,1564,1565,1568,1569,1571,1572,1574,1575,1577],{},[263,1566,1567],{},"For ultra-high resolution",": FLUX.2 ",[1398,1570,1400],{},", FLUX.2 ",[1398,1573,1418],{},", and FLUX.2 ",[1398,1576,1435],{}," support custom resolutions up to 14142px, making them ideal for large-format prints and detailed compositions.",[20,1579,1581],{"id":1580},"aspect-ratios","Aspect Ratios",[10,1583,1584],{},"Melies supports 9 aspect ratios so you can generate images in the exact format you need.",[33,1586,1587,1598],{},[36,1588,1589],{},[39,1590,1591,1593,1596],{},[42,1592,950],{},[42,1594,1595],{},"Name",[42,1597,887],{},[67,1599,1600,1611,1622,1633,1644,1655,1666,1677,1688],{},[39,1601,1602,1605,1608],{},[72,1603,1604],{},"Auto",[72,1606,1607],{},"Automatic",[72,1609,1610],{},"Let the model choose the optimal ratio",[39,1612,1613,1616,1619],{},[72,1614,1615],{},"21:9",[72,1617,1618],{},"Cinematic",[72,1620,1621],{},"Widescreen film compositions, epic landscapes",[39,1623,1624,1627,1630],{},[72,1625,1626],{},"16:9",[72,1628,1629],{},"Landscape",[72,1631,1632],{},"Standard widescreen, YouTube thumbnails",[39,1634,1635,1638,1641],{},[72,1636,1637],{},"4:3",[72,1639,1640],{},"Classic",[72,1642,1643],{},"Traditional film format, TV-style compositions",[39,1645,1646,1649,1652],{},[72,1647,1648],{},"3:2",[72,1650,1651],{},"Photo",[72,1653,1654],{},"Standard photography ratio, natural framing",[39,1656,1657,1660,1663],{},[72,1658,1659],{},"1:1",[72,1661,1662],{},"Square",[72,1664,1665],{},"Social media posts, album covers, portraits",[39,1667,1668,1671,1674],{},[72,1669,1670],{},"2:3",[72,1672,1673],{},"Photo Portrait",[72,1675,1676],{},"Vertical photography, character portraits",[39,1678,1679,1682,1685],{},[72,1680,1681],{},"3:4",[72,1683,1684],{},"Portrait",[72,1686,1687],{},"Vertical compositions, mobile-friendly content",[39,1689,1690,1693,1696],{},[72,1691,1692],{},"9:16",[72,1694,1695],{},"Portrait/Mobile",[72,1697,1698],{},"Instagram Stories, TikTok, vertical video frames",[10,1700,1701],{},"Select the aspect ratio before generating your image. Not all models support all ratios - the interface will show you which options are available for your selected model.",[20,1703,1705],{"id":1704},"resolution-settings","Resolution Settings",[10,1707,1708],{},"Resolution options vary by model.",[328,1710,1711,1717,1723,1737],{},[306,1712,1713,1716],{},[263,1714,1715],{},"Google models"," (Nano Banana 2, Nano Banana Pro): Support resolutions from 0.5K up to 4K. Higher resolutions cost more credits. Nano Banana 2 ranges from 8 credits at 0.5K to 40 credits at 4K.",[306,1718,1719,1722],{},[263,1720,1721],{},"ByteDance models"," (Seedream 4 Edit, Seedream 4.5): Support Auto, 2K, and 4K output.",[306,1724,1725,1728,1729,1731,1732,1731,1734,1736],{},[263,1726,1727],{},"FLUX.2 series"," (",[1398,1730,1400],{},", ",[1398,1733,1418],{},[1398,1735,1435],{},"): Support custom resolutions up to 14142px on any side.",[306,1738,1739,1742],{},[263,1740,1741],{},"Standard Flux models",": Generate at standard resolutions tied to the selected aspect ratio.",[10,1744,1745],{},"For most filmmaking workflows, generating at standard resolution and then upscaling the best results is the most credit-efficient approach.",[20,1747,1749],{"id":1748},"using-reference-images","Using Reference Images",[10,1751,1752],{},"Reference images let you guide the AI with a visual starting point instead of relying solely on text prompts.",[28,1754,1756],{"id":1755},"how-to-use-a-reference-image","How to Use a Reference Image",[303,1758,1759,1763,1766,1769,1772],{},[306,1760,308,1761],{},[14,1762,373],{"href":372},[306,1764,1765],{},"Click the reference image upload area",[306,1767,1768],{},"Upload your reference image (a photo, sketch, or any visual)",[306,1770,1771],{},"Write your prompt describing what you want to generate",[306,1773,1774],{},"The AI will use your reference as a visual guide alongside your text prompt",[10,1776,1777],{},"Reference images work well for:",[328,1779,1780,1786,1792,1798],{},[306,1781,1782,1785],{},[263,1783,1784],{},"Maintaining character consistency"," across multiple shots in your film",[306,1787,1788,1791],{},[263,1789,1790],{},"Converting sketches to polished images"," by uploading rough drawings",[306,1793,1794,1797],{},[263,1795,1796],{},"Style transfer"," by uploading an image with a visual style you want to replicate",[306,1799,1800,1803],{},[263,1801,1802],{},"Scene variations"," by using an existing frame and asking for changes",[20,1805,1807],{"id":1806},"extracting-a-subject-from-an-image","Extracting a Subject from an Image",[10,1809,1810],{},"Melies can extract the main subject from any image, which is useful for maintaining character consistency across scenes.",[303,1812,1813,1816,1819,1822],{},[306,1814,1815],{},"Generate or upload an image with a character or subject",[306,1817,1818],{},"Use the extract subject feature",[306,1820,1821],{},"The AI isolates the subject from the background",[306,1823,1824],{},"Use the extracted subject as a reference for new generations",[10,1826,1827],{},"This is particularly powerful when you have established a character design and want to place them in different settings, lighting conditions, or scenes throughout your project.",[20,1829,1831],{"id":1830},"generating-variations","Generating Variations",[10,1833,1834],{},"One of the most powerful features in the Melies image generator is the variation system. After generating an image, you can create targeted variations across 11 categories without rewriting your prompt.",[28,1836,1838],{"id":1837},"camera-angles-9-options","Camera Angles (9 options)",[10,1840,1841],{},"Change the perspective of your shot while keeping the same subject and scene.",[328,1843,1844,1847,1850,1853,1856,1859,1862,1865,1868],{},[306,1845,1846],{},"Eye level - Standard natural perspective",[306,1848,1849],{},"High angle - Looking down on the subject",[306,1851,1852],{},"Low angle - Looking up, adds power and drama",[306,1854,1855],{},"Overhead/Top-down - Bird's eye view",[306,1857,1858],{},"Ground level - Camera at ground height",[306,1860,1861],{},"Dutch tilt - Angled frame for tension and unease",[306,1863,1864],{},"Over-the-shoulder - Common in dialogue scenes",[306,1866,1867],{},"Profile view - Side-on perspective",[306,1869,1870],{},"Three-quarter - Angled between frontal and profile",[28,1872,1874],{"id":1873},"shot-sizes-9-options","Shot Sizes (9 options)",[10,1876,1877],{},"Adjust framing from extreme close-ups to wide establishing shots.",[328,1879,1880,1883,1886,1889,1892,1895,1898,1901,1904],{},[306,1881,1882],{},"Tighter (zoom in) - Move closer to the current framing",[306,1884,1885],{},"Wider (zoom out) - Pull back from the current framing",[306,1887,1888],{},"Extreme close-up - Eyes, mouth, or small details only",[306,1890,1891],{},"Close-up - Face fills the frame",[306,1893,1894],{},"Medium close-up - Head and shoulders",[306,1896,1897],{},"Medium shot - Waist up",[306,1899,1900],{},"Cowboy shot - Mid-thigh up (classic Western framing)",[306,1902,1903],{},"Full body - Entire figure visible",[306,1905,1906],{},"Wide shot - Subject with surrounding environment",[28,1908,1910],{"id":1909},"expressions-23-options","Expressions (23 options)",[10,1912,1913],{},"Change a character's facial expression to match the emotion of your scene.",[10,1915,1916,1919],{},[263,1917,1918],{},"Base expressions",": Neutral, Natural smile, Laughing, Serious, Surprised, Sad, Angry, Crying",[10,1921,1922,1925],{},[263,1923,1924],{},"Influencer pack",": Wink, Tongue out, Duck face, Blowing a kiss, Smize, Biting lip",[10,1927,1928,1931],{},[263,1929,1930],{},"Cinema pack",": Villain smirk, Thousand-yard stare, Seductive, Horrified, Wonder/Awe, Triumphant, Exhausted, Nervous, Contemplative, Grieving, Skeptical",[28,1933,1935],{"id":1934},"lighting-10-options","Lighting (10 options)",[10,1937,1938],{},"Transform the mood of your image through lighting changes.",[328,1940,1941,1944,1947,1950,1953,1956,1959,1962,1965,1968],{},[306,1942,1943],{},"Soft/Diffused - Even, flattering light with minimal shadows",[306,1945,1946],{},"Hard/Dramatic - Strong directional light with deep shadows",[306,1948,1949],{},"Golden hour - Warm, low-angle sunlight",[306,1951,1952],{},"Blue hour/Dusk - Cool twilight tones",[306,1954,1955],{},"Low-key Noir - Dark with selective highlights",[306,1957,1958],{},"High-key Bright - Bright and airy with minimal shadows",[306,1960,1961],{},"Rembrandt - Classic portrait lighting with triangular highlight",[306,1963,1964],{},"Backlit/Rim - Light from behind the subject creating a glow outline",[306,1966,1967],{},"Candlelight - Warm, flickering, intimate",[306,1969,1970],{},"Neon glow - Vibrant colored artificial light",[28,1972,1974],{"id":1973},"time-of-day-7-options","Time of Day (7 options)",[10,1976,1977],{},"Shift the time setting while keeping your composition intact.",[10,1979,1980],{},"Dawn, Sunrise, Morning, Midday, Golden hour, Dusk/Blue hour, Night",[28,1982,1984],{"id":1983},"weather-7-options","Weather (7 options)",[10,1986,1987],{},"Add atmospheric conditions to your scene.",[10,1989,1990],{},"Clear/Sunny, Overcast, Foggy, Misty, Rainy, Stormy, Snowy",[28,1992,1994],{"id":1993},"color-grading-10-options","Color Grading (10 options)",[10,1996,1997],{},"Apply cinematic color grades to change the visual tone.",[328,1999,2000,2003,2006,2009,2012,2015,2018,2021,2024,2027],{},[306,2001,2002],{},"Natural - True-to-life colors",[306,2004,2005],{},"Teal & Orange - Hollywood blockbuster look",[306,2007,2008],{},"Monochrome - Black and white",[306,2010,2011],{},"Warm amber - Cozy, nostalgic warmth",[306,2013,2014],{},"Cool blue - Clinical, cold, or futuristic",[306,2016,2017],{},"Filmic/Faded - Vintage film stock feel",[306,2019,2020],{},"Bleach bypass - Desaturated with high contrast",[306,2022,2023],{},"Sepia - Classic aged photograph look",[306,2025,2026],{},"Desaturated - Muted, understated palette",[306,2028,2029],{},"Hyper-saturated - Bold, vivid colors",[28,2031,2033],{"id":2032},"mood-12-options","Mood (12 options)",[10,2035,2036],{},"Set the emotional atmosphere of the entire image.",[10,2038,2039],{},"Romantic, Mysterious, Tense/Thriller, Ethereal/Dreamy, Gritty/Raw, Melancholic, Epic/Heroic, Nostalgic, Eerie/Unsettling, Joyful/Uplifting, Serene/Peaceful, Dark/Foreboding",[28,2041,2043],{"id":2042},"art-styles-17-options","Art Styles (17 options)",[10,2045,2046],{},"Transform your image into a completely different artistic style.",[10,2048,2049],{},"Cinematic film still, Hollywood blockbuster, Arthouse/Indie, Classic film noir, Neo-noir, Documentary, Anime, Studio Ghibli, Makoto Shinkai, Comic book, Pixar/3D animation, Oil painting, Watercolor, Charcoal sketch, Concept art, Synthwave/80s, Fashion editorial",[28,2051,2053],{"id":2052},"eras-12-options","Eras (12 options)",[10,2055,2056],{},"Place your scene in a different historical or future time period.",[10,2058,2059],{},"Victorian/1880s, 1920s Art Deco, 1950s post-war, 1960s/Mod, 1970s/New Hollywood, 1980s/Neon, 1990s/Grunge, Early 2000s, Modern contemporary, Near future, Dystopian future, Medieval/Fantasy",[28,2061,950],{"id":2062},"aspect-ratio",[10,2064,2065],{},"Switch between any of the 9 available aspect ratios to reframe your composition for different platforms or creative needs.",[20,2067,2069],{"id":2068},"right-click-context-menu","Right-Click Context Menu",[10,2071,2072],{},"Every generated image has a context menu (right-click or long-press on mobile) that gives you quick access to powerful features.",[328,2074,2075,2081,2087,2093,2098],{},[306,2076,2077,2080],{},[263,2078,2079],{},"Generate variations"," - Open the variation panel for any of the 11 categories listed above",[306,2082,2083,2086],{},[263,2084,2085],{},"Extract subject"," - Isolate the main subject for use as a reference",[306,2088,2089,2092],{},[263,2090,2091],{},"Upscale"," - Increase resolution up to 4K",[306,2094,2095,2097],{},[263,2096,850],{}," - Set this image as the reference for your next generation",[306,2099,2100,2103],{},[263,2101,2102],{},"Download"," - Save the image to your device",[10,2105,2106],{},"This context menu makes it fast to iterate on your best results without navigating away from your workspace.",[20,2108,2110],{"id":2109},"upscaling","Upscaling",[10,2112,2113],{},"Melies includes built-in image upscaling to take any generated image to higher resolution.",[28,2115,2117],{"id":2116},"how-to-upscale","How to Upscale",[303,2119,2120,2123,2126,2129],{},[306,2121,2122],{},"Generate an image using any model",[306,2124,2125],{},"Right-click the image and select \"Upscale\"",[306,2127,2128],{},"Choose your target resolution (up to 4K)",[306,2130,2131],{},"The upscaler enhances detail while increasing resolution",[28,2133,2135],{"id":2134},"when-to-upscale","When to Upscale",[328,2137,2138,2144,2150],{},[306,2139,2140,2143],{},[263,2141,2142],{},"Final production images"," that need to be print-quality or used in high-resolution video",[306,2145,2146,2149],{},[263,2147,2148],{},"Images from fast/cheap models"," like Flux Schnell - generate cheaply, then upscale only the best results",[306,2151,2152,2155],{},[263,2153,2154],{},"Close-up shots"," where you need sharp detail on faces and textures",[10,2157,2158],{},"Upscaling is a cost-effective workflow: generate many options with affordable models, pick your favorites, then upscale only the winners.",[20,2160,981],{"id":980},[10,2162,2163],{},"Here is a quick reference for credit costs across all image models.",[33,2165,2166,2178],{},[36,2167,2168],{},[39,2169,2170,2173,2176],{},[42,2171,2172],{},"Budget Level",[42,2174,2175],{},"Models",[42,2177,56],{},[67,2179,2180,2189,2200,2210,2223,2236],{},[39,2181,2182,2185,2187],{},[72,2183,2184],{},"Economy",[72,2186,1516],{},[72,2188,1526],{},[39,2190,2191,2194,2198],{},[72,2192,2193],{},"Budget",[72,2195,1396,2196],{},[1398,2197,1452],{},[72,2199,1462],{},[39,2201,2202,2205,2208],{},[72,2203,2204],{},"Mid-range",[72,2206,2207],{},"Nano Banana, Seedream 4 Edit, Seedream 4.5, Flux Dev",[72,2209,1306],{},[39,2211,2212,2215,2221],{},[72,2213,2214],{},"Standard",[72,2216,2217,2218,2220],{},"Nano Banana 2, Nano Banana Pro, Grok Imagine, FLUX.2 ",[1398,2219,1435],{},", Flux Pro",[72,2222,1289],{},[39,2224,2225,2228,2234],{},[72,2226,2227],{},"Premium",[72,2229,2230,2231,2233],{},"Flux Kontext, FLUX.2 ",[1398,2232,1418],{},", Flux Pro Ultra",[72,2235,1389],{},[39,2237,2238,2241,2246],{},[72,2239,2240],{},"Ultra",[72,2242,2243,2244],{},"Flux Kontext Max, FLUX.2 ",[1398,2245,1400],{},[72,2247,1372],{},[10,2249,2250],{},"Nano Banana 2 has variable pricing based on resolution: 8 credits at 0.5K up to 40 credits at 4K.",[10,2252,1078,2253,2255],{},[14,2254,1082],{"href":1081}," page for current credit package options.",[20,2257,2259],{"id":2258},"tips-for-great-results","Tips for Great Results",[10,2261,2262,2265],{},[263,2263,2264],{},"Write detailed prompts",": The more specific your prompt, the better. Include details about the subject, setting, lighting, mood, and camera angle.",[10,2267,2268,2271,2272,2274],{},[263,2269,2270],{},"Use the right model for the job",": Do not spend 25 credits on a concept sketch. Start with Flux Schnell or FLUX.2 ",[1398,2273,1452],{}," to explore ideas, then switch to premium models for final images.",[10,2276,2277,2280],{},[263,2278,2279],{},"Leverage variations",": Instead of rewriting prompts from scratch, generate one good base image and use variations to explore different angles, lighting, and moods.",[10,2282,2283,2286],{},[263,2284,2285],{},"Maintain character consistency",": Use the extract subject feature and reference images to keep your characters looking the same across multiple scenes.",[10,2288,2289,2292],{},[263,2290,2291],{},"Match your aspect ratio to your project",": If you are making a film, stick with 21:9 or 16:9. For social media content, use 9:16 or 1:1.",[10,2294,2295,2296,2298,2299,2301],{},"Ready to start creating? Open the ",[14,2297,373],{"href":372}," and bring your vision to life. Check out the ",[14,2300,17],{"href":1188}," to turn your best images into moving shots.",{"title":1129,"searchDepth":1130,"depth":1130,"links":2303},[2304,2308,2309,2310,2313,2314,2327,2328,2332,2333],{"id":22,"depth":1130,"text":23,"children":2305},[2306,2307],{"id":30,"depth":1135,"text":31},{"id":258,"depth":1135,"text":259},{"id":1580,"depth":1130,"text":1581},{"id":1704,"depth":1130,"text":1705},{"id":1748,"depth":1130,"text":1749,"children":2311},[2312],{"id":1755,"depth":1135,"text":1756},{"id":1806,"depth":1130,"text":1807},{"id":1830,"depth":1130,"text":1831,"children":2315},[2316,2317,2318,2319,2320,2321,2322,2323,2324,2325,2326],{"id":1837,"depth":1135,"text":1838},{"id":1873,"depth":1135,"text":1874},{"id":1909,"depth":1135,"text":1910},{"id":1934,"depth":1135,"text":1935},{"id":1973,"depth":1135,"text":1974},{"id":1983,"depth":1135,"text":1984},{"id":1993,"depth":1135,"text":1994},{"id":2032,"depth":1135,"text":2033},{"id":2042,"depth":1135,"text":2043},{"id":2052,"depth":1135,"text":2053},{"id":2062,"depth":1135,"text":950},{"id":2068,"depth":1130,"text":2069},{"id":2109,"depth":1130,"text":2110,"children":2329},[2330,2331],{"id":2116,"depth":1135,"text":2117},{"id":2134,"depth":1135,"text":2135},{"id":980,"depth":1130,"text":981},{"id":2258,"depth":1130,"text":2259},"Learn how to use Melies AI Image Generator with 16 models, 9 aspect ratios, reference images, variations, and upscaling. Full guide for filmmakers.","/img/blog/ai-image-generator-overview.webp",{"sitemap":2337},{"loc":1126,"lastmod":1181,"changefreq":1189,"priority":1190},"/en/docs/ai-image-generator",[2340,2343,2346,2349,2352,2355],{"question":2341,"answer":2342},"How many AI image models does Melies offer?","Melies offers 16 AI image generation models from providers including Google, ByteDance, xAI, and Black Forest Labs. Models range from fast drafts (Flux Schnell at 2 credits) to premium quality (Flux Kontext Max and FLUX.2 max at 25 credits).",{"question":2344,"answer":2345},"What aspect ratios are available for AI image generation?","Melies supports 9 aspect ratios: Auto, 21:9 (Cinematic), 16:9 (Landscape), 4:3 (Classic), 3:2 (Photo), 1:1 (Square), 2:3 (Photo Portrait), 3:4 (Portrait), and 9:16 (Portrait/Mobile).",{"question":2347,"answer":2348},"Can I use reference images with the AI image generator?","Yes. You can upload a reference image and the AI will use it as a visual guide when generating new images. You can also extract a subject from an existing image to maintain character consistency across multiple shots.",{"question":2350,"answer":2351},"How do image variations work in Melies?","Melies offers 11 variation categories including camera angles, shot sizes, expressions, lighting, time of day, weather, color grading, mood, art style, era, and aspect ratio. Right-click any generated image to quickly create variations without rewriting your prompt.",{"question":2353,"answer":2354},"What is the cheapest AI image model on Melies?","Flux Schnell is the most affordable model at just 2 credits per image. It is the fastest model available and works well for quick drafts and concept exploration.",{"question":2356,"answer":2357},"Can I upscale AI-generated images on Melies?","Yes. Melies includes a built-in upscaling tool that lets you increase the resolution of any generated image up to 4K, giving you print-quality results from any model.",{"title":1223,"description":2334},"en/docs/ai-image-generator",[2361,1217,1218,2362],"ai image generator","image generation","qrErBjBQCFmzpZgR7YC9t7pEFnvgHtr94ZTEm8Zd02U"]