{"object":"model_capability.list","data":[{"id":"minimax-h3-mini","available":true,"status":"available","description":"MiniMax H3 Mini video generation model.","model":"mini","tier":"mini","precision":"int8","sampling_steps":4,"modes":["t2va","i2va","fl2va","l2va","ref2va"],"resolutions":["480p","720p"],"aspect_ratios":["16:9","9:16","1:1","4:3","3:4","21:9","9:21","4:5","5:4"],"canvas":{"480p":{"16:9":"864x480","9:16":"480x864","1:1":"640x640","4:3":"736x544","3:4":"544x736","21:9":"960x416","9:21":"416x992","4:5":"576x704","5:4":"704x576"},"720p":{"16:9":"1280x736","9:16":"736x1280","1:1":"960x960","4:3":"1120x832","3:4":"832x1120","21:9":"1504x640","9:21":"640x1472","4:5":"864x1088","5:4":"1088x864"}},"input":{"content_types":["application/json","multipart/form-data"],"parameters":[{"name":"model","type":"string","required":true,"values":["minimax-h3-mini"],"description":"MiniMax H3 model type identifier."},{"name":"mode","type":"enum","required":true,"values":["t2va","i2va","fl2va","l2va","ref2va"],"description":"Generation mode selected independently from the model tier."},{"name":"resolution","type":"enum","required":true,"values":["480p","720p"],"description":"Output resolution tier."},{"name":"prompt","type":"string","required":true,"max_chars":30000,"description":"Text prompt describing the scene, motion, and audio."},{"name":"prompt_optimization","type":"boolean","required":false,"default":false,"description":"Rewrite prompt into the official H3 structure while preserving user intent."},{"name":"seconds","type":"integer|string","required":true,"enum":[5,6,7,8,9,10,11,12,13,14,15],"values":[5,6,7,8,9,10,11,12,13,14,15],"min":5,"max":15,"description":"Requested output duration. Allowed whole-second values are 5 through 15; billing uses output seconds."},{"name":"aspect_ratio","type":"enum","required":false,"values":["16:9","9:16","1:1","4:3","3:4","21:9","9:21","4:5","5:4"],"default":"16:9","description":"Output frame ratio. 16:9 horizontal and 9:16 vertical are canonical; 1:1, 4:3/3:4, 21:9/9:21, and 4:5/5:4 are also supported. Pricing is independent of aspect ratio."},{"type":"file|url","required":false,"max_items":9,"max_file_size_bytes":20971520,"description":"Keyframe images for i2va/fl2va/l2va, or reference images for ref2va.","asset_object":{"url":"http(s) URL when not uploading a local file"},"name":"images"},{"type":"file|url","required":false,"max_items":3,"max_file_size_bytes":52428800,"description":"Reference audio clips for ref2va.","asset_object":{"url":"http(s) URL when not uploading a local file"},"name":"audios"}],"requirements_by_mode":{"t2va":{"images":0,"videos":0,"audios":0},"i2va":{"images":1,"videos":0,"audios":0},"fl2va":{"images":2,"videos":0,"audios":0},"l2va":{"images":1,"videos":0,"audios":0},"ref2va":{"required_any_of":["images","audios"],"max_total":12}},"prompt_guidance_by_mode":{"t2va":{"passthrough":"verbatim when prompt_optimization is false","structured_prompt_required":false,"optimization":{"parameter":"prompt_optimization","type":"boolean","default":false,"behavior":"When true, the service rewrites the prompt into the official H3 structure while preserving user intent, dialogue, lyrics, and visible text."},"content_review":{"always_applied":true,"stages":["original","optimized_when_present"],"rejection_status":422,"rejection_code":"prompt_review_rejected","unavailable_status":503,"unavailable_code":"prompt_review_unavailable","rewrites_prompt":false},"language_rule":"Write structural prose in the user's input language. Chinese input produces Chinese scene, action, camera, sound, and music descriptions; fixed H3 field names and language tags remain protocol syntax. Preserve dialogue, lyrics, and visible scene text verbatim.","syntax":{"shot":"[Shot N] only for an actual edited shot boundary","timed_shot":"[Shot N] At MM:SS.mmm, ... for an actual later cut","timed_event":"At MM:SS.mmm, ... for a timed event inside the current shot","continuity":"Use one [Shot 1] for continuous camera/scene action; expression, dialogue, and action phases do not create new shots.","speaker":"(S1) or (S1,S2)","dialogue":"<d>[Language] exact dialogue</d>","cross_cut_dialogue":"<scenetrans>","truncated_dialogue":"<cutoff>","visible_text":"Preserve visible text verbatim inside English double quotes.","voiceover":"Use 'says in an off-screen voiceover' and state that the corresponding on-screen character's lips remain completely closed."},"camera_guidance":{"motion_types":["Zoom In","Zoom Out","Push In","Pull Out","Pan Left","Pan Right","Truck Left","Truck Right","Tilt Up","Tilt Down","Pedestal Up","Pedestal Down","Arc Shot","Tracking Shot","Static Shot","Shake Slightly","Shake Strongly","POV","Roll Clockwise","Roll Counterclockwise"],"amplitude":["with small amplitude","with large amplitude"],"speed":["at slow speed","at fast speed"],"rule":"Write camera movement as a natural action within its shot."},"format":"base","section_order":["integrated_multimodal_description","overall_soundscape","non_diegetic_music"],"keyframe_instruction":null,"media_labels":{"images":[]}},"i2va":{"passthrough":"verbatim when prompt_optimization is false","structured_prompt_required":false,"optimization":{"parameter":"prompt_optimization","type":"boolean","default":false,"behavior":"When true, the service rewrites the prompt into the official H3 structure while preserving user intent, dialogue, lyrics, and visible text."},"content_review":{"always_applied":true,"stages":["original","optimized_when_present"],"rejection_status":422,"rejection_code":"prompt_review_rejected","unavailable_status":503,"unavailable_code":"prompt_review_unavailable","rewrites_prompt":false},"language_rule":"Write structural prose in the user's input language. Chinese input produces Chinese scene, action, camera, sound, and music descriptions; fixed H3 field names and language tags remain protocol syntax. Preserve dialogue, lyrics, and visible scene text verbatim.","syntax":{"shot":"[Shot N] only for an actual edited shot boundary","timed_shot":"[Shot N] At MM:SS.mmm, ... for an actual later cut","timed_event":"At MM:SS.mmm, ... for a timed event inside the current shot","continuity":"Use one [Shot 1] for continuous camera/scene action; expression, dialogue, and action phases do not create new shots.","speaker":"(S1) or (S1,S2)","dialogue":"<d>[Language] exact dialogue</d>","cross_cut_dialogue":"<scenetrans>","truncated_dialogue":"<cutoff>","visible_text":"Preserve visible text verbatim inside English double quotes.","voiceover":"Use 'says in an off-screen voiceover' and state that the corresponding on-screen character's lips remain completely closed."},"camera_guidance":{"motion_types":["Zoom In","Zoom Out","Push In","Pull Out","Pan Left","Pan Right","Truck Left","Truck Right","Tilt Up","Tilt Down","Pedestal Up","Pedestal Down","Arc Shot","Tracking Shot","Static Shot","Shake Slightly","Shake Strongly","POV","Roll Clockwise","Roll Counterclockwise"],"amplitude":["with small amplitude","with large amplitude"],"speed":["at slow speed","at fast speed"],"rule":"Write camera movement as a natural action within its shot."},"format":"base","section_order":["integrated_multimodal_description","overall_soundscape","non_diegetic_music"],"keyframe_instruction":"For the target video, at 0.00 seconds into the target video, <Picture 1> (from [Shot 1]) is fully referenced.","media_labels":{"images":["<Picture 1>"]}},"fl2va":{"passthrough":"verbatim when prompt_optimization is false","structured_prompt_required":false,"optimization":{"parameter":"prompt_optimization","type":"boolean","default":false,"behavior":"When true, the service rewrites the prompt into the official H3 structure while preserving user intent, dialogue, lyrics, and visible text."},"content_review":{"always_applied":true,"stages":["original","optimized_when_present"],"rejection_status":422,"rejection_code":"prompt_review_rejected","unavailable_status":503,"unavailable_code":"prompt_review_unavailable","rewrites_prompt":false},"language_rule":"Write structural prose in the user's input language. Chinese input produces Chinese scene, action, camera, sound, and music descriptions; fixed H3 field names and language tags remain protocol syntax. Preserve dialogue, lyrics, and visible scene text verbatim.","syntax":{"shot":"[Shot N] only for an actual edited shot boundary","timed_shot":"[Shot N] At MM:SS.mmm, ... for an actual later cut","timed_event":"At MM:SS.mmm, ... for a timed event inside the current shot","continuity":"Use one [Shot 1] for continuous camera/scene action; expression, dialogue, and action phases do not create new shots.","speaker":"(S1) or (S1,S2)","dialogue":"<d>[Language] exact dialogue</d>","cross_cut_dialogue":"<scenetrans>","truncated_dialogue":"<cutoff>","visible_text":"Preserve visible text verbatim inside English double quotes.","voiceover":"Use 'says in an off-screen voiceover' and state that the corresponding on-screen character's lips remain completely closed."},"camera_guidance":{"motion_types":["Zoom In","Zoom Out","Push In","Pull Out","Pan Left","Pan Right","Truck Left","Truck Right","Tilt Up","Tilt Down","Pedestal Up","Pedestal Down","Arc Shot","Tracking Shot","Static Shot","Shake Slightly","Shake Strongly","POV","Roll Clockwise","Roll Counterclockwise"],"amplitude":["with small amplitude","with large amplitude"],"speed":["at slow speed","at fast speed"],"rule":"Write camera movement as a natural action within its shot."},"format":"base","section_order":["integrated_multimodal_description","overall_soundscape","non_diegetic_music"],"keyframe_instruction":"How the reference pictures align with the target video - Picture 1 (from Shot 1) aligns with the 0.00-second mark of the target video; Picture 2 (from Shot N) aligns with the S.SS-second mark of the target video.","media_labels":{"images":["<Picture 1>","<Picture 2>"]}},"l2va":{"passthrough":"verbatim when prompt_optimization is false","structured_prompt_required":false,"optimization":{"parameter":"prompt_optimization","type":"boolean","default":false,"behavior":"When true, the service rewrites the prompt into the official H3 structure while preserving user intent, dialogue, lyrics, and visible text."},"content_review":{"always_applied":true,"stages":["original","optimized_when_present"],"rejection_status":422,"rejection_code":"prompt_review_rejected","unavailable_status":503,"unavailable_code":"prompt_review_unavailable","rewrites_prompt":false},"language_rule":"Write structural prose in the user's input language. Chinese input produces Chinese scene, action, camera, sound, and music descriptions; fixed H3 field names and language tags remain protocol syntax. Preserve dialogue, lyrics, and visible scene text verbatim.","syntax":{"shot":"[Shot N] only for an actual edited shot boundary","timed_shot":"[Shot N] At MM:SS.mmm, ... for an actual later cut","timed_event":"At MM:SS.mmm, ... for a timed event inside the current shot","continuity":"Use one [Shot 1] for continuous camera/scene action; expression, dialogue, and action phases do not create new shots.","speaker":"(S1) or (S1,S2)","dialogue":"<d>[Language] exact dialogue</d>","cross_cut_dialogue":"<scenetrans>","truncated_dialogue":"<cutoff>","visible_text":"Preserve visible text verbatim inside English double quotes.","voiceover":"Use 'says in an off-screen voiceover' and state that the corresponding on-screen character's lips remain completely closed."},"camera_guidance":{"motion_types":["Zoom In","Zoom Out","Push In","Pull Out","Pan Left","Pan Right","Truck Left","Truck Right","Tilt Up","Tilt Down","Pedestal Up","Pedestal Down","Arc Shot","Tracking Shot","Static Shot","Shake Slightly","Shake Strongly","POV","Roll Clockwise","Roll Counterclockwise"],"amplitude":["with small amplitude","with large amplitude"],"speed":["at slow speed","at fast speed"],"rule":"Write camera movement as a natural action within its shot."},"format":"base","section_order":["integrated_multimodal_description","overall_soundscape","non_diegetic_music"],"keyframe_instruction":"How the reference pictures align with the target video - <Picture 1> (from [Shot N]) aligns with the S.SS-second mark of the target video.","media_labels":{"images":["<Picture 1>"]}},"ref2va":{"passthrough":"verbatim when prompt_optimization is false","structured_prompt_required":false,"optimization":{"parameter":"prompt_optimization","type":"boolean","default":false,"behavior":"When true, the service rewrites the prompt into the official H3 structure while preserving user intent, dialogue, lyrics, and visible text."},"content_review":{"always_applied":true,"stages":["original","optimized_when_present"],"rejection_status":422,"rejection_code":"prompt_review_rejected","unavailable_status":503,"unavailable_code":"prompt_review_unavailable","rewrites_prompt":false},"language_rule":"Write all prose in all six Ref2VA sections in English, regardless of the user's input language. Preserve only dialogue, lyrics, and visible scene text verbatim in their original language.","syntax":{"shot":"[Shot N] only for an actual edited shot boundary","timed_shot":"[Shot N] At MM:SS.mmm, ... for an actual later cut","timed_event":"At MM:SS.mmm, ... for a timed event inside the current shot","continuity":"Use one [Shot 1] for continuous camera/scene action; expression, dialogue, and action phases do not create new shots.","speaker":"(S1) or (S1,S2)","dialogue":"<d>[Language] exact dialogue</d>","cross_cut_dialogue":"<scenetrans>","truncated_dialogue":"<cutoff>","visible_text":"Preserve visible text verbatim inside English double quotes.","voiceover":"Use 'says in an off-screen voiceover' and state that the corresponding on-screen character's lips remain completely closed.","subject":"<Subject N>","picture":"<Picture N>","audio":"<Audio N>"},"camera_guidance":{"motion_types":["Zoom In","Zoom Out","Push In","Pull Out","Pan Left","Pan Right","Truck Left","Truck Right","Tilt Up","Tilt Down","Pedestal Up","Pedestal Down","Arc Shot","Tracking Shot","Static Shot","Shake Slightly","Shake Strongly","POV","Roll Clockwise","Roll Counterclockwise"],"amplitude":["with small amplitude","with large amplitude"],"speed":["at slow speed","at fast speed"],"rule":"Write camera movement as a natural action within its shot."},"format":"full_reference","section_order":["subject_definitions","summary","retention_analysis","detailed_description","overall_soundscape","non_diegetic_music"],"media_labels":{"images":"<Picture N> in images upload order","audios":"<Audio N> in audios upload order","numbering":"Picture, Audio, and Subject labels use independent counters."},"summary_task_types":["keyframe completion","reference generation","audio reuse","audio reference"],"retention_markers":{"visual":["fully_preserved","partially_preserved","attribute_transfer","weak_reference"],"audio":["fully_copy","partially_copy","reference","weak_reference"]},"detailed_description_budget":{"target_words_per_second":35,"maximum_words":500,"priority":"reference-role and timeline fidelity over word count"}}},"limits":{"prompt_max_chars":30000,"seconds_min":5,"seconds_max":15,"seconds_values":[5,6,7,8,9,10,11,12,13,14,15],"max_images":9,"max_videos":0,"max_audios":3,"max_media":12,"max_asset_size_bytes":52428800,"max_image_size_bytes":20971520,"max_video_size_bytes":0,"max_audio_size_bytes":52428800,"max_upload_total_size_bytes":125829120,"max_multipart_body_size_bytes":134217728},"media_preprocessing":{"images":{"automatic":true,"trigger_size_bytes":8388608,"max_long_edge":2048,"output_format":"image/jpeg"}}},"output":{"content_type":"video/mp4","fps":24,"status_endpoint":"/v1/videos/{id}","content_endpoint":"/v1/videos/{id}/content"},"billing":{"currency":"CNY","unit":"output_second","aspect_ratio_independent":true,"price_per_second_cny":{"480p":0.04,"720p":0.05}}},{"id":"minimax-h3-fast","available":true,"status":"available","description":"MiniMax H3 Fast video generation model.","model":"fast","tier":"fast","precision":"int8","sampling_steps":8,"modes":["t2va","i2va","fl2va","l2va","ref2va"],"resolutions":["480p","720p"],"aspect_ratios":["16:9","9:16","1:1","4:3","3:4","21:9","9:21","4:5","5:4"],"canvas":{"480p":{"16:9":"864x480","9:16":"480x864","1:1":"640x640","4:3":"736x544","3:4":"544x736","21:9":"960x416","9:21":"416x992","4:5":"576x704","5:4":"704x576"},"720p":{"16:9":"1280x736","9:16":"736x1280","1:1":"960x960","4:3":"1120x832","3:4":"832x1120","21:9":"1504x640","9:21":"640x1472","4:5":"864x1088","5:4":"1088x864"}},"input":{"content_types":["application/json","multipart/form-data"],"parameters":[{"name":"model","type":"string","required":true,"values":["minimax-h3-fast"],"description":"MiniMax H3 model type identifier."},{"name":"mode","type":"enum","required":true,"values":["t2va","i2va","fl2va","l2va","ref2va"],"description":"Generation mode selected independently from the model tier."},{"name":"resolution","type":"enum","required":true,"values":["480p","720p"],"description":"Output resolution tier."},{"name":"prompt","type":"string","required":true,"max_chars":30000,"description":"Text prompt describing the scene, motion, and audio."},{"name":"prompt_optimization","type":"boolean","required":false,"default":false,"description":"Rewrite prompt into the official H3 structure while preserving user intent."},{"name":"seconds","type":"integer|string","required":true,"enum":[5,6,7,8,9,10,11,12,13,14,15],"values":[5,6,7,8,9,10,11,12,13,14,15],"min":5,"max":15,"description":"Requested output duration. Allowed whole-second values are 5 through 15; billing uses output seconds."},{"name":"aspect_ratio","type":"enum","required":false,"values":["16:9","9:16","1:1","4:3","3:4","21:9","9:21","4:5","5:4"],"default":"16:9","description":"Output frame ratio. 16:9 horizontal and 9:16 vertical are canonical; 1:1, 4:3/3:4, 21:9/9:21, and 4:5/5:4 are also supported. Pricing is independent of aspect ratio."},{"type":"file|url","required":false,"max_items":9,"max_file_size_bytes":20971520,"description":"Keyframe images for i2va/fl2va/l2va, or reference images for ref2va.","asset_object":{"url":"http(s) URL when not uploading a local file"},"name":"images"},{"type":"file|url","required":false,"max_items":3,"max_file_size_bytes":52428800,"description":"Reference audio clips for ref2va.","asset_object":{"url":"http(s) URL when not uploading a local file"},"name":"audios"}],"requirements_by_mode":{"t2va":{"images":0,"videos":0,"audios":0},"i2va":{"images":1,"videos":0,"audios":0},"fl2va":{"images":2,"videos":0,"audios":0},"l2va":{"images":1,"videos":0,"audios":0},"ref2va":{"required_any_of":["images","audios"],"max_total":12}},"prompt_guidance_by_mode":{"t2va":{"passthrough":"verbatim when prompt_optimization is false","structured_prompt_required":false,"optimization":{"parameter":"prompt_optimization","type":"boolean","default":false,"behavior":"When true, the service rewrites the prompt into the official H3 structure while preserving user intent, dialogue, lyrics, and visible text."},"content_review":{"always_applied":true,"stages":["original","optimized_when_present"],"rejection_status":422,"rejection_code":"prompt_review_rejected","unavailable_status":503,"unavailable_code":"prompt_review_unavailable","rewrites_prompt":false},"language_rule":"Write structural prose in the user's input language. Chinese input produces Chinese scene, action, camera, sound, and music descriptions; fixed H3 field names and language tags remain protocol syntax. Preserve dialogue, lyrics, and visible scene text verbatim.","syntax":{"shot":"[Shot N] only for an actual edited shot boundary","timed_shot":"[Shot N] At MM:SS.mmm, ... for an actual later cut","timed_event":"At MM:SS.mmm, ... for a timed event inside the current shot","continuity":"Use one [Shot 1] for continuous camera/scene action; expression, dialogue, and action phases do not create new shots.","speaker":"(S1) or (S1,S2)","dialogue":"<d>[Language] exact dialogue</d>","cross_cut_dialogue":"<scenetrans>","truncated_dialogue":"<cutoff>","visible_text":"Preserve visible text verbatim inside English double quotes.","voiceover":"Use 'says in an off-screen voiceover' and state that the corresponding on-screen character's lips remain completely closed."},"camera_guidance":{"motion_types":["Zoom In","Zoom Out","Push In","Pull Out","Pan Left","Pan Right","Truck Left","Truck Right","Tilt Up","Tilt Down","Pedestal Up","Pedestal Down","Arc Shot","Tracking Shot","Static Shot","Shake Slightly","Shake Strongly","POV","Roll Clockwise","Roll Counterclockwise"],"amplitude":["with small amplitude","with large amplitude"],"speed":["at slow speed","at fast speed"],"rule":"Write camera movement as a natural action within its shot."},"format":"base","section_order":["integrated_multimodal_description","overall_soundscape","non_diegetic_music"],"keyframe_instruction":null,"media_labels":{"images":[]}},"i2va":{"passthrough":"verbatim when prompt_optimization is false","structured_prompt_required":false,"optimization":{"parameter":"prompt_optimization","type":"boolean","default":false,"behavior":"When true, the service rewrites the prompt into the official H3 structure while preserving user intent, dialogue, lyrics, and visible text."},"content_review":{"always_applied":true,"stages":["original","optimized_when_present"],"rejection_status":422,"rejection_code":"prompt_review_rejected","unavailable_status":503,"unavailable_code":"prompt_review_unavailable","rewrites_prompt":false},"language_rule":"Write structural prose in the user's input language. Chinese input produces Chinese scene, action, camera, sound, and music descriptions; fixed H3 field names and language tags remain protocol syntax. Preserve dialogue, lyrics, and visible scene text verbatim.","syntax":{"shot":"[Shot N] only for an actual edited shot boundary","timed_shot":"[Shot N] At MM:SS.mmm, ... for an actual later cut","timed_event":"At MM:SS.mmm, ... for a timed event inside the current shot","continuity":"Use one [Shot 1] for continuous camera/scene action; expression, dialogue, and action phases do not create new shots.","speaker":"(S1) or (S1,S2)","dialogue":"<d>[Language] exact dialogue</d>","cross_cut_dialogue":"<scenetrans>","truncated_dialogue":"<cutoff>","visible_text":"Preserve visible text verbatim inside English double quotes.","voiceover":"Use 'says in an off-screen voiceover' and state that the corresponding on-screen character's lips remain completely closed."},"camera_guidance":{"motion_types":["Zoom In","Zoom Out","Push In","Pull Out","Pan Left","Pan Right","Truck Left","Truck Right","Tilt Up","Tilt Down","Pedestal Up","Pedestal Down","Arc Shot","Tracking Shot","Static Shot","Shake Slightly","Shake Strongly","POV","Roll Clockwise","Roll Counterclockwise"],"amplitude":["with small amplitude","with large amplitude"],"speed":["at slow speed","at fast speed"],"rule":"Write camera movement as a natural action within its shot."},"format":"base","section_order":["integrated_multimodal_description","overall_soundscape","non_diegetic_music"],"keyframe_instruction":"For the target video, at 0.00 seconds into the target video, <Picture 1> (from [Shot 1]) is fully referenced.","media_labels":{"images":["<Picture 1>"]}},"fl2va":{"passthrough":"verbatim when prompt_optimization is false","structured_prompt_required":false,"optimization":{"parameter":"prompt_optimization","type":"boolean","default":false,"behavior":"When true, the service rewrites the prompt into the official H3 structure while preserving user intent, dialogue, lyrics, and visible text."},"content_review":{"always_applied":true,"stages":["original","optimized_when_present"],"rejection_status":422,"rejection_code":"prompt_review_rejected","unavailable_status":503,"unavailable_code":"prompt_review_unavailable","rewrites_prompt":false},"language_rule":"Write structural prose in the user's input language. Chinese input produces Chinese scene, action, camera, sound, and music descriptions; fixed H3 field names and language tags remain protocol syntax. Preserve dialogue, lyrics, and visible scene text verbatim.","syntax":{"shot":"[Shot N] only for an actual edited shot boundary","timed_shot":"[Shot N] At MM:SS.mmm, ... for an actual later cut","timed_event":"At MM:SS.mmm, ... for a timed event inside the current shot","continuity":"Use one [Shot 1] for continuous camera/scene action; expression, dialogue, and action phases do not create new shots.","speaker":"(S1) or (S1,S2)","dialogue":"<d>[Language] exact dialogue</d>","cross_cut_dialogue":"<scenetrans>","truncated_dialogue":"<cutoff>","visible_text":"Preserve visible text verbatim inside English double quotes.","voiceover":"Use 'says in an off-screen voiceover' and state that the corresponding on-screen character's lips remain completely closed."},"camera_guidance":{"motion_types":["Zoom In","Zoom Out","Push In","Pull Out","Pan Left","Pan Right","Truck Left","Truck Right","Tilt Up","Tilt Down","Pedestal Up","Pedestal Down","Arc Shot","Tracking Shot","Static Shot","Shake Slightly","Shake Strongly","POV","Roll Clockwise","Roll Counterclockwise"],"amplitude":["with small amplitude","with large amplitude"],"speed":["at slow speed","at fast speed"],"rule":"Write camera movement as a natural action within its shot."},"format":"base","section_order":["integrated_multimodal_description","overall_soundscape","non_diegetic_music"],"keyframe_instruction":"How the reference pictures align with the target video - Picture 1 (from Shot 1) aligns with the 0.00-second mark of the target video; Picture 2 (from Shot N) aligns with the S.SS-second mark of the target video.","media_labels":{"images":["<Picture 1>","<Picture 2>"]}},"l2va":{"passthrough":"verbatim when prompt_optimization is false","structured_prompt_required":false,"optimization":{"parameter":"prompt_optimization","type":"boolean","default":false,"behavior":"When true, the service rewrites the prompt into the official H3 structure while preserving user intent, dialogue, lyrics, and visible text."},"content_review":{"always_applied":true,"stages":["original","optimized_when_present"],"rejection_status":422,"rejection_code":"prompt_review_rejected","unavailable_status":503,"unavailable_code":"prompt_review_unavailable","rewrites_prompt":false},"language_rule":"Write structural prose in the user's input language. Chinese input produces Chinese scene, action, camera, sound, and music descriptions; fixed H3 field names and language tags remain protocol syntax. Preserve dialogue, lyrics, and visible scene text verbatim.","syntax":{"shot":"[Shot N] only for an actual edited shot boundary","timed_shot":"[Shot N] At MM:SS.mmm, ... for an actual later cut","timed_event":"At MM:SS.mmm, ... for a timed event inside the current shot","continuity":"Use one [Shot 1] for continuous camera/scene action; expression, dialogue, and action phases do not create new shots.","speaker":"(S1) or (S1,S2)","dialogue":"<d>[Language] exact dialogue</d>","cross_cut_dialogue":"<scenetrans>","truncated_dialogue":"<cutoff>","visible_text":"Preserve visible text verbatim inside English double quotes.","voiceover":"Use 'says in an off-screen voiceover' and state that the corresponding on-screen character's lips remain completely closed."},"camera_guidance":{"motion_types":["Zoom In","Zoom Out","Push In","Pull Out","Pan Left","Pan Right","Truck Left","Truck Right","Tilt Up","Tilt Down","Pedestal Up","Pedestal Down","Arc Shot","Tracking Shot","Static Shot","Shake Slightly","Shake Strongly","POV","Roll Clockwise","Roll Counterclockwise"],"amplitude":["with small amplitude","with large amplitude"],"speed":["at slow speed","at fast speed"],"rule":"Write camera movement as a natural action within its shot."},"format":"base","section_order":["integrated_multimodal_description","overall_soundscape","non_diegetic_music"],"keyframe_instruction":"How the reference pictures align with the target video - <Picture 1> (from [Shot N]) aligns with the S.SS-second mark of the target video.","media_labels":{"images":["<Picture 1>"]}},"ref2va":{"passthrough":"verbatim when prompt_optimization is false","structured_prompt_required":false,"optimization":{"parameter":"prompt_optimization","type":"boolean","default":false,"behavior":"When true, the service rewrites the prompt into the official H3 structure while preserving user intent, dialogue, lyrics, and visible text."},"content_review":{"always_applied":true,"stages":["original","optimized_when_present"],"rejection_status":422,"rejection_code":"prompt_review_rejected","unavailable_status":503,"unavailable_code":"prompt_review_unavailable","rewrites_prompt":false},"language_rule":"Write all prose in all six Ref2VA sections in English, regardless of the user's input language. Preserve only dialogue, lyrics, and visible scene text verbatim in their original language.","syntax":{"shot":"[Shot N] only for an actual edited shot boundary","timed_shot":"[Shot N] At MM:SS.mmm, ... for an actual later cut","timed_event":"At MM:SS.mmm, ... for a timed event inside the current shot","continuity":"Use one [Shot 1] for continuous camera/scene action; expression, dialogue, and action phases do not create new shots.","speaker":"(S1) or (S1,S2)","dialogue":"<d>[Language] exact dialogue</d>","cross_cut_dialogue":"<scenetrans>","truncated_dialogue":"<cutoff>","visible_text":"Preserve visible text verbatim inside English double quotes.","voiceover":"Use 'says in an off-screen voiceover' and state that the corresponding on-screen character's lips remain completely closed.","subject":"<Subject N>","picture":"<Picture N>","audio":"<Audio N>"},"camera_guidance":{"motion_types":["Zoom In","Zoom Out","Push In","Pull Out","Pan Left","Pan Right","Truck Left","Truck Right","Tilt Up","Tilt Down","Pedestal Up","Pedestal Down","Arc Shot","Tracking Shot","Static Shot","Shake Slightly","Shake Strongly","POV","Roll Clockwise","Roll Counterclockwise"],"amplitude":["with small amplitude","with large amplitude"],"speed":["at slow speed","at fast speed"],"rule":"Write camera movement as a natural action within its shot."},"format":"full_reference","section_order":["subject_definitions","summary","retention_analysis","detailed_description","overall_soundscape","non_diegetic_music"],"media_labels":{"images":"<Picture N> in images upload order","audios":"<Audio N> in audios upload order","numbering":"Picture, Audio, and Subject labels use independent counters."},"summary_task_types":["keyframe completion","reference generation","audio reuse","audio reference"],"retention_markers":{"visual":["fully_preserved","partially_preserved","attribute_transfer","weak_reference"],"audio":["fully_copy","partially_copy","reference","weak_reference"]},"detailed_description_budget":{"target_words_per_second":35,"maximum_words":500,"priority":"reference-role and timeline fidelity over word count"}}},"limits":{"prompt_max_chars":30000,"seconds_min":5,"seconds_max":15,"seconds_values":[5,6,7,8,9,10,11,12,13,14,15],"max_images":9,"max_videos":0,"max_audios":3,"max_media":12,"max_asset_size_bytes":52428800,"max_image_size_bytes":20971520,"max_video_size_bytes":0,"max_audio_size_bytes":52428800,"max_upload_total_size_bytes":125829120,"max_multipart_body_size_bytes":134217728},"media_preprocessing":{"images":{"automatic":true,"trigger_size_bytes":8388608,"max_long_edge":2048,"output_format":"image/jpeg"}}},"output":{"content_type":"video/mp4","fps":24,"status_endpoint":"/v1/videos/{id}","content_endpoint":"/v1/videos/{id}/content"},"billing":{"currency":"CNY","unit":"output_second","aspect_ratio_independent":true,"price_per_second_cny":{"480p":0.05,"720p":0.06}}},{"id":"minimax-h3-base","available":true,"status":"available","description":"MiniMax H3 Base video generation model.","model":"base","tier":"base","precision":"int8","sampling_steps":16,"modes":["t2va","i2va","fl2va","l2va","ref2va"],"resolutions":["480p","720p"],"aspect_ratios":["16:9","9:16","1:1","4:3","3:4","21:9","9:21","4:5","5:4"],"canvas":{"480p":{"16:9":"864x480","9:16":"480x864","1:1":"640x640","4:3":"736x544","3:4":"544x736","21:9":"960x416","9:21":"416x992","4:5":"576x704","5:4":"704x576"},"720p":{"16:9":"1280x736","9:16":"736x1280","1:1":"960x960","4:3":"1120x832","3:4":"832x1120","21:9":"1504x640","9:21":"640x1472","4:5":"864x1088","5:4":"1088x864"}},"input":{"content_types":["application/json","multipart/form-data"],"parameters":[{"name":"model","type":"string","required":true,"values":["minimax-h3-base"],"description":"MiniMax H3 model type identifier."},{"name":"mode","type":"enum","required":true,"values":["t2va","i2va","fl2va","l2va","ref2va"],"description":"Generation mode selected independently from the model tier."},{"name":"resolution","type":"enum","required":true,"values":["480p","720p"],"description":"Output resolution tier."},{"name":"prompt","type":"string","required":true,"max_chars":30000,"description":"Text prompt describing the scene, motion, and audio."},{"name":"prompt_optimization","type":"boolean","required":false,"default":false,"description":"Rewrite prompt into the official H3 structure while preserving user intent."},{"name":"seconds","type":"integer|string","required":true,"enum":[5,6,7,8,9,10,11,12,13,14,15],"values":[5,6,7,8,9,10,11,12,13,14,15],"min":5,"max":15,"description":"Requested output duration. Allowed whole-second values are 5 through 15; billing uses output seconds."},{"name":"aspect_ratio","type":"enum","required":false,"values":["16:9","9:16","1:1","4:3","3:4","21:9","9:21","4:5","5:4"],"default":"16:9","description":"Output frame ratio. 16:9 horizontal and 9:16 vertical are canonical; 1:1, 4:3/3:4, 21:9/9:21, and 4:5/5:4 are also supported. Pricing is independent of aspect ratio."},{"type":"file|url","required":false,"max_items":9,"max_file_size_bytes":20971520,"description":"Keyframe images for i2va/fl2va/l2va, or reference images for ref2va.","asset_object":{"url":"http(s) URL when not uploading a local file"},"name":"images"},{"type":"file|url","required":false,"max_items":3,"max_file_size_bytes":52428800,"description":"Reference audio clips for ref2va.","asset_object":{"url":"http(s) URL when not uploading a local file"},"name":"audios"}],"requirements_by_mode":{"t2va":{"images":0,"videos":0,"audios":0},"i2va":{"images":1,"videos":0,"audios":0},"fl2va":{"images":2,"videos":0,"audios":0},"l2va":{"images":1,"videos":0,"audios":0},"ref2va":{"required_any_of":["images","audios"],"max_total":12}},"prompt_guidance_by_mode":{"t2va":{"passthrough":"verbatim when prompt_optimization is false","structured_prompt_required":false,"optimization":{"parameter":"prompt_optimization","type":"boolean","default":false,"behavior":"When true, the service rewrites the prompt into the official H3 structure while preserving user intent, dialogue, lyrics, and visible text."},"content_review":{"always_applied":true,"stages":["original","optimized_when_present"],"rejection_status":422,"rejection_code":"prompt_review_rejected","unavailable_status":503,"unavailable_code":"prompt_review_unavailable","rewrites_prompt":false},"language_rule":"Write structural prose in the user's input language. Chinese input produces Chinese scene, action, camera, sound, and music descriptions; fixed H3 field names and language tags remain protocol syntax. Preserve dialogue, lyrics, and visible scene text verbatim.","syntax":{"shot":"[Shot N] only for an actual edited shot boundary","timed_shot":"[Shot N] At MM:SS.mmm, ... for an actual later cut","timed_event":"At MM:SS.mmm, ... for a timed event inside the current shot","continuity":"Use one [Shot 1] for continuous camera/scene action; expression, dialogue, and action phases do not create new shots.","speaker":"(S1) or (S1,S2)","dialogue":"<d>[Language] exact dialogue</d>","cross_cut_dialogue":"<scenetrans>","truncated_dialogue":"<cutoff>","visible_text":"Preserve visible text verbatim inside English double quotes.","voiceover":"Use 'says in an off-screen voiceover' and state that the corresponding on-screen character's lips remain completely closed."},"camera_guidance":{"motion_types":["Zoom In","Zoom Out","Push In","Pull Out","Pan Left","Pan Right","Truck Left","Truck Right","Tilt Up","Tilt Down","Pedestal Up","Pedestal Down","Arc Shot","Tracking Shot","Static Shot","Shake Slightly","Shake Strongly","POV","Roll Clockwise","Roll Counterclockwise"],"amplitude":["with small amplitude","with large amplitude"],"speed":["at slow speed","at fast speed"],"rule":"Write camera movement as a natural action within its shot."},"format":"base","section_order":["integrated_multimodal_description","overall_soundscape","non_diegetic_music"],"keyframe_instruction":null,"media_labels":{"images":[]}},"i2va":{"passthrough":"verbatim when prompt_optimization is false","structured_prompt_required":false,"optimization":{"parameter":"prompt_optimization","type":"boolean","default":false,"behavior":"When true, the service rewrites the prompt into the official H3 structure while preserving user intent, dialogue, lyrics, and visible text."},"content_review":{"always_applied":true,"stages":["original","optimized_when_present"],"rejection_status":422,"rejection_code":"prompt_review_rejected","unavailable_status":503,"unavailable_code":"prompt_review_unavailable","rewrites_prompt":false},"language_rule":"Write structural prose in the user's input language. Chinese input produces Chinese scene, action, camera, sound, and music descriptions; fixed H3 field names and language tags remain protocol syntax. Preserve dialogue, lyrics, and visible scene text verbatim.","syntax":{"shot":"[Shot N] only for an actual edited shot boundary","timed_shot":"[Shot N] At MM:SS.mmm, ... for an actual later cut","timed_event":"At MM:SS.mmm, ... for a timed event inside the current shot","continuity":"Use one [Shot 1] for continuous camera/scene action; expression, dialogue, and action phases do not create new shots.","speaker":"(S1) or (S1,S2)","dialogue":"<d>[Language] exact dialogue</d>","cross_cut_dialogue":"<scenetrans>","truncated_dialogue":"<cutoff>","visible_text":"Preserve visible text verbatim inside English double quotes.","voiceover":"Use 'says in an off-screen voiceover' and state that the corresponding on-screen character's lips remain completely closed."},"camera_guidance":{"motion_types":["Zoom In","Zoom Out","Push In","Pull Out","Pan Left","Pan Right","Truck Left","Truck Right","Tilt Up","Tilt Down","Pedestal Up","Pedestal Down","Arc Shot","Tracking Shot","Static Shot","Shake Slightly","Shake Strongly","POV","Roll Clockwise","Roll Counterclockwise"],"amplitude":["with small amplitude","with large amplitude"],"speed":["at slow speed","at fast speed"],"rule":"Write camera movement as a natural action within its shot."},"format":"base","section_order":["integrated_multimodal_description","overall_soundscape","non_diegetic_music"],"keyframe_instruction":"For the target video, at 0.00 seconds into the target video, <Picture 1> (from [Shot 1]) is fully referenced.","media_labels":{"images":["<Picture 1>"]}},"fl2va":{"passthrough":"verbatim when prompt_optimization is false","structured_prompt_required":false,"optimization":{"parameter":"prompt_optimization","type":"boolean","default":false,"behavior":"When true, the service rewrites the prompt into the official H3 structure while preserving user intent, dialogue, lyrics, and visible text."},"content_review":{"always_applied":true,"stages":["original","optimized_when_present"],"rejection_status":422,"rejection_code":"prompt_review_rejected","unavailable_status":503,"unavailable_code":"prompt_review_unavailable","rewrites_prompt":false},"language_rule":"Write structural prose in the user's input language. Chinese input produces Chinese scene, action, camera, sound, and music descriptions; fixed H3 field names and language tags remain protocol syntax. Preserve dialogue, lyrics, and visible scene text verbatim.","syntax":{"shot":"[Shot N] only for an actual edited shot boundary","timed_shot":"[Shot N] At MM:SS.mmm, ... for an actual later cut","timed_event":"At MM:SS.mmm, ... for a timed event inside the current shot","continuity":"Use one [Shot 1] for continuous camera/scene action; expression, dialogue, and action phases do not create new shots.","speaker":"(S1) or (S1,S2)","dialogue":"<d>[Language] exact dialogue</d>","cross_cut_dialogue":"<scenetrans>","truncated_dialogue":"<cutoff>","visible_text":"Preserve visible text verbatim inside English double quotes.","voiceover":"Use 'says in an off-screen voiceover' and state that the corresponding on-screen character's lips remain completely closed."},"camera_guidance":{"motion_types":["Zoom In","Zoom Out","Push In","Pull Out","Pan Left","Pan Right","Truck Left","Truck Right","Tilt Up","Tilt Down","Pedestal Up","Pedestal Down","Arc Shot","Tracking Shot","Static Shot","Shake Slightly","Shake Strongly","POV","Roll Clockwise","Roll Counterclockwise"],"amplitude":["with small amplitude","with large amplitude"],"speed":["at slow speed","at fast speed"],"rule":"Write camera movement as a natural action within its shot."},"format":"base","section_order":["integrated_multimodal_description","overall_soundscape","non_diegetic_music"],"keyframe_instruction":"How the reference pictures align with the target video - Picture 1 (from Shot 1) aligns with the 0.00-second mark of the target video; Picture 2 (from Shot N) aligns with the S.SS-second mark of the target video.","media_labels":{"images":["<Picture 1>","<Picture 2>"]}},"l2va":{"passthrough":"verbatim when prompt_optimization is false","structured_prompt_required":false,"optimization":{"parameter":"prompt_optimization","type":"boolean","default":false,"behavior":"When true, the service rewrites the prompt into the official H3 structure while preserving user intent, dialogue, lyrics, and visible text."},"content_review":{"always_applied":true,"stages":["original","optimized_when_present"],"rejection_status":422,"rejection_code":"prompt_review_rejected","unavailable_status":503,"unavailable_code":"prompt_review_unavailable","rewrites_prompt":false},"language_rule":"Write structural prose in the user's input language. Chinese input produces Chinese scene, action, camera, sound, and music descriptions; fixed H3 field names and language tags remain protocol syntax. Preserve dialogue, lyrics, and visible scene text verbatim.","syntax":{"shot":"[Shot N] only for an actual edited shot boundary","timed_shot":"[Shot N] At MM:SS.mmm, ... for an actual later cut","timed_event":"At MM:SS.mmm, ... for a timed event inside the current shot","continuity":"Use one [Shot 1] for continuous camera/scene action; expression, dialogue, and action phases do not create new shots.","speaker":"(S1) or (S1,S2)","dialogue":"<d>[Language] exact dialogue</d>","cross_cut_dialogue":"<scenetrans>","truncated_dialogue":"<cutoff>","visible_text":"Preserve visible text verbatim inside English double quotes.","voiceover":"Use 'says in an off-screen voiceover' and state that the corresponding on-screen character's lips remain completely closed."},"camera_guidance":{"motion_types":["Zoom In","Zoom Out","Push In","Pull Out","Pan Left","Pan Right","Truck Left","Truck Right","Tilt Up","Tilt Down","Pedestal Up","Pedestal Down","Arc Shot","Tracking Shot","Static Shot","Shake Slightly","Shake Strongly","POV","Roll Clockwise","Roll Counterclockwise"],"amplitude":["with small amplitude","with large amplitude"],"speed":["at slow speed","at fast speed"],"rule":"Write camera movement as a natural action within its shot."},"format":"base","section_order":["integrated_multimodal_description","overall_soundscape","non_diegetic_music"],"keyframe_instruction":"How the reference pictures align with the target video - <Picture 1> (from [Shot N]) aligns with the S.SS-second mark of the target video.","media_labels":{"images":["<Picture 1>"]}},"ref2va":{"passthrough":"verbatim when prompt_optimization is false","structured_prompt_required":false,"optimization":{"parameter":"prompt_optimization","type":"boolean","default":false,"behavior":"When true, the service rewrites the prompt into the official H3 structure while preserving user intent, dialogue, lyrics, and visible text."},"content_review":{"always_applied":true,"stages":["original","optimized_when_present"],"rejection_status":422,"rejection_code":"prompt_review_rejected","unavailable_status":503,"unavailable_code":"prompt_review_unavailable","rewrites_prompt":false},"language_rule":"Write all prose in all six Ref2VA sections in English, regardless of the user's input language. Preserve only dialogue, lyrics, and visible scene text verbatim in their original language.","syntax":{"shot":"[Shot N] only for an actual edited shot boundary","timed_shot":"[Shot N] At MM:SS.mmm, ... for an actual later cut","timed_event":"At MM:SS.mmm, ... for a timed event inside the current shot","continuity":"Use one [Shot 1] for continuous camera/scene action; expression, dialogue, and action phases do not create new shots.","speaker":"(S1) or (S1,S2)","dialogue":"<d>[Language] exact dialogue</d>","cross_cut_dialogue":"<scenetrans>","truncated_dialogue":"<cutoff>","visible_text":"Preserve visible text verbatim inside English double quotes.","voiceover":"Use 'says in an off-screen voiceover' and state that the corresponding on-screen character's lips remain completely closed.","subject":"<Subject N>","picture":"<Picture N>","audio":"<Audio N>"},"camera_guidance":{"motion_types":["Zoom In","Zoom Out","Push In","Pull Out","Pan Left","Pan Right","Truck Left","Truck Right","Tilt Up","Tilt Down","Pedestal Up","Pedestal Down","Arc Shot","Tracking Shot","Static Shot","Shake Slightly","Shake Strongly","POV","Roll Clockwise","Roll Counterclockwise"],"amplitude":["with small amplitude","with large amplitude"],"speed":["at slow speed","at fast speed"],"rule":"Write camera movement as a natural action within its shot."},"format":"full_reference","section_order":["subject_definitions","summary","retention_analysis","detailed_description","overall_soundscape","non_diegetic_music"],"media_labels":{"images":"<Picture N> in images upload order","audios":"<Audio N> in audios upload order","numbering":"Picture, Audio, and Subject labels use independent counters."},"summary_task_types":["keyframe completion","reference generation","audio reuse","audio reference"],"retention_markers":{"visual":["fully_preserved","partially_preserved","attribute_transfer","weak_reference"],"audio":["fully_copy","partially_copy","reference","weak_reference"]},"detailed_description_budget":{"target_words_per_second":35,"maximum_words":500,"priority":"reference-role and timeline fidelity over word count"}}},"limits":{"prompt_max_chars":30000,"seconds_min":5,"seconds_max":15,"seconds_values":[5,6,7,8,9,10,11,12,13,14,15],"max_images":9,"max_videos":0,"max_audios":3,"max_media":12,"max_asset_size_bytes":52428800,"max_image_size_bytes":20971520,"max_video_size_bytes":0,"max_audio_size_bytes":52428800,"max_upload_total_size_bytes":125829120,"max_multipart_body_size_bytes":134217728},"media_preprocessing":{"images":{"automatic":true,"trigger_size_bytes":8388608,"max_long_edge":2048,"output_format":"image/jpeg"}}},"output":{"content_type":"video/mp4","fps":24,"status_endpoint":"/v1/videos/{id}","content_endpoint":"/v1/videos/{id}/content"},"billing":{"currency":"CNY","unit":"output_second","aspect_ratio_independent":true,"price_per_second_cny":{"480p":0.06,"720p":0.07}}},{"id":"minimax-h3-pro","available":false,"status":"coming_soon","description":"MiniMax H3 Pro video generation model.","model":"pro","tier":"pro","precision":"bf16","sampling_steps":25,"modes":["t2va","i2va","fl2va","l2va","ref2va"],"resolutions":["480p","720p"],"aspect_ratios":["16:9","9:16","1:1","4:3","3:4","21:9","9:21","4:5","5:4"],"canvas":{"480p":{"16:9":"864x480","9:16":"480x864","1:1":"640x640","4:3":"736x544","3:4":"544x736","21:9":"960x416","9:21":"416x992","4:5":"576x704","5:4":"704x576"},"720p":{"16:9":"1280x736","9:16":"736x1280","1:1":"960x960","4:3":"1120x832","3:4":"832x1120","21:9":"1504x640","9:21":"640x1472","4:5":"864x1088","5:4":"1088x864"}},"input":{"content_types":["application/json","multipart/form-data"],"parameters":[{"name":"model","type":"string","required":true,"values":["minimax-h3-pro"],"description":"MiniMax H3 model type identifier."},{"name":"mode","type":"enum","required":true,"values":["t2va","i2va","fl2va","l2va","ref2va"],"description":"Generation mode selected independently from the model tier."},{"name":"resolution","type":"enum","required":true,"values":["480p","720p"],"description":"Output resolution tier."},{"name":"prompt","type":"string","required":true,"max_chars":30000,"description":"Text prompt describing the scene, motion, and audio."},{"name":"prompt_optimization","type":"boolean","required":false,"default":false,"description":"Rewrite prompt into the official H3 structure while preserving user intent."},{"name":"seconds","type":"integer|string","required":true,"enum":[5,6,7,8,9,10,11,12,13,14,15],"values":[5,6,7,8,9,10,11,12,13,14,15],"min":5,"max":15,"description":"Requested output duration. Allowed whole-second values are 5 through 15; billing uses output seconds."},{"name":"aspect_ratio","type":"enum","required":false,"values":["16:9","9:16","1:1","4:3","3:4","21:9","9:21","4:5","5:4"],"default":"16:9","description":"Output frame ratio. 16:9 horizontal and 9:16 vertical are canonical; 1:1, 4:3/3:4, 21:9/9:21, and 4:5/5:4 are also supported. Pricing is independent of aspect ratio."},{"type":"file|url","required":false,"max_items":9,"max_file_size_bytes":20971520,"description":"Keyframe images for i2va/fl2va/l2va, or reference images for ref2va.","asset_object":{"url":"http(s) URL when not uploading a local file"},"name":"images"},{"type":"file|url","required":false,"max_items":3,"max_file_size_bytes":52428800,"description":"Reference audio clips for ref2va.","asset_object":{"url":"http(s) URL when not uploading a local file"},"name":"audios"}],"requirements_by_mode":{"t2va":{"images":0,"videos":0,"audios":0},"i2va":{"images":1,"videos":0,"audios":0},"fl2va":{"images":2,"videos":0,"audios":0},"l2va":{"images":1,"videos":0,"audios":0},"ref2va":{"required_any_of":["images","audios"],"max_total":12}},"prompt_guidance_by_mode":{"t2va":{"passthrough":"verbatim when prompt_optimization is false","structured_prompt_required":false,"optimization":{"parameter":"prompt_optimization","type":"boolean","default":false,"behavior":"When true, the service rewrites the prompt into the official H3 structure while preserving user intent, dialogue, lyrics, and visible text."},"content_review":{"always_applied":true,"stages":["original","optimized_when_present"],"rejection_status":422,"rejection_code":"prompt_review_rejected","unavailable_status":503,"unavailable_code":"prompt_review_unavailable","rewrites_prompt":false},"language_rule":"Write structural prose in the user's input language. Chinese input produces Chinese scene, action, camera, sound, and music descriptions; fixed H3 field names and language tags remain protocol syntax. Preserve dialogue, lyrics, and visible scene text verbatim.","syntax":{"shot":"[Shot N] only for an actual edited shot boundary","timed_shot":"[Shot N] At MM:SS.mmm, ... for an actual later cut","timed_event":"At MM:SS.mmm, ... for a timed event inside the current shot","continuity":"Use one [Shot 1] for continuous camera/scene action; expression, dialogue, and action phases do not create new shots.","speaker":"(S1) or (S1,S2)","dialogue":"<d>[Language] exact dialogue</d>","cross_cut_dialogue":"<scenetrans>","truncated_dialogue":"<cutoff>","visible_text":"Preserve visible text verbatim inside English double quotes.","voiceover":"Use 'says in an off-screen voiceover' and state that the corresponding on-screen character's lips remain completely closed."},"camera_guidance":{"motion_types":["Zoom In","Zoom Out","Push In","Pull Out","Pan Left","Pan Right","Truck Left","Truck Right","Tilt Up","Tilt Down","Pedestal Up","Pedestal Down","Arc Shot","Tracking Shot","Static Shot","Shake Slightly","Shake Strongly","POV","Roll Clockwise","Roll Counterclockwise"],"amplitude":["with small amplitude","with large amplitude"],"speed":["at slow speed","at fast speed"],"rule":"Write camera movement as a natural action within its shot."},"format":"base","section_order":["integrated_multimodal_description","overall_soundscape","non_diegetic_music"],"keyframe_instruction":null,"media_labels":{"images":[]}},"i2va":{"passthrough":"verbatim when prompt_optimization is false","structured_prompt_required":false,"optimization":{"parameter":"prompt_optimization","type":"boolean","default":false,"behavior":"When true, the service rewrites the prompt into the official H3 structure while preserving user intent, dialogue, lyrics, and visible text."},"content_review":{"always_applied":true,"stages":["original","optimized_when_present"],"rejection_status":422,"rejection_code":"prompt_review_rejected","unavailable_status":503,"unavailable_code":"prompt_review_unavailable","rewrites_prompt":false},"language_rule":"Write structural prose in the user's input language. Chinese input produces Chinese scene, action, camera, sound, and music descriptions; fixed H3 field names and language tags remain protocol syntax. Preserve dialogue, lyrics, and visible scene text verbatim.","syntax":{"shot":"[Shot N] only for an actual edited shot boundary","timed_shot":"[Shot N] At MM:SS.mmm, ... for an actual later cut","timed_event":"At MM:SS.mmm, ... for a timed event inside the current shot","continuity":"Use one [Shot 1] for continuous camera/scene action; expression, dialogue, and action phases do not create new shots.","speaker":"(S1) or (S1,S2)","dialogue":"<d>[Language] exact dialogue</d>","cross_cut_dialogue":"<scenetrans>","truncated_dialogue":"<cutoff>","visible_text":"Preserve visible text verbatim inside English double quotes.","voiceover":"Use 'says in an off-screen voiceover' and state that the corresponding on-screen character's lips remain completely closed."},"camera_guidance":{"motion_types":["Zoom In","Zoom Out","Push In","Pull Out","Pan Left","Pan Right","Truck Left","Truck Right","Tilt Up","Tilt Down","Pedestal Up","Pedestal Down","Arc Shot","Tracking Shot","Static Shot","Shake Slightly","Shake Strongly","POV","Roll Clockwise","Roll Counterclockwise"],"amplitude":["with small amplitude","with large amplitude"],"speed":["at slow speed","at fast speed"],"rule":"Write camera movement as a natural action within its shot."},"format":"base","section_order":["integrated_multimodal_description","overall_soundscape","non_diegetic_music"],"keyframe_instruction":"For the target video, at 0.00 seconds into the target video, <Picture 1> (from [Shot 1]) is fully referenced.","media_labels":{"images":["<Picture 1>"]}},"fl2va":{"passthrough":"verbatim when prompt_optimization is false","structured_prompt_required":false,"optimization":{"parameter":"prompt_optimization","type":"boolean","default":false,"behavior":"When true, the service rewrites the prompt into the official H3 structure while preserving user intent, dialogue, lyrics, and visible text."},"content_review":{"always_applied":true,"stages":["original","optimized_when_present"],"rejection_status":422,"rejection_code":"prompt_review_rejected","unavailable_status":503,"unavailable_code":"prompt_review_unavailable","rewrites_prompt":false},"language_rule":"Write structural prose in the user's input language. Chinese input produces Chinese scene, action, camera, sound, and music descriptions; fixed H3 field names and language tags remain protocol syntax. Preserve dialogue, lyrics, and visible scene text verbatim.","syntax":{"shot":"[Shot N] only for an actual edited shot boundary","timed_shot":"[Shot N] At MM:SS.mmm, ... for an actual later cut","timed_event":"At MM:SS.mmm, ... for a timed event inside the current shot","continuity":"Use one [Shot 1] for continuous camera/scene action; expression, dialogue, and action phases do not create new shots.","speaker":"(S1) or (S1,S2)","dialogue":"<d>[Language] exact dialogue</d>","cross_cut_dialogue":"<scenetrans>","truncated_dialogue":"<cutoff>","visible_text":"Preserve visible text verbatim inside English double quotes.","voiceover":"Use 'says in an off-screen voiceover' and state that the corresponding on-screen character's lips remain completely closed."},"camera_guidance":{"motion_types":["Zoom In","Zoom Out","Push In","Pull Out","Pan Left","Pan Right","Truck Left","Truck Right","Tilt Up","Tilt Down","Pedestal Up","Pedestal Down","Arc Shot","Tracking Shot","Static Shot","Shake Slightly","Shake Strongly","POV","Roll Clockwise","Roll Counterclockwise"],"amplitude":["with small amplitude","with large amplitude"],"speed":["at slow speed","at fast speed"],"rule":"Write camera movement as a natural action within its shot."},"format":"base","section_order":["integrated_multimodal_description","overall_soundscape","non_diegetic_music"],"keyframe_instruction":"How the reference pictures align with the target video - Picture 1 (from Shot 1) aligns with the 0.00-second mark of the target video; Picture 2 (from Shot N) aligns with the S.SS-second mark of the target video.","media_labels":{"images":["<Picture 1>","<Picture 2>"]}},"l2va":{"passthrough":"verbatim when prompt_optimization is false","structured_prompt_required":false,"optimization":{"parameter":"prompt_optimization","type":"boolean","default":false,"behavior":"When true, the service rewrites the prompt into the official H3 structure while preserving user intent, dialogue, lyrics, and visible text."},"content_review":{"always_applied":true,"stages":["original","optimized_when_present"],"rejection_status":422,"rejection_code":"prompt_review_rejected","unavailable_status":503,"unavailable_code":"prompt_review_unavailable","rewrites_prompt":false},"language_rule":"Write structural prose in the user's input language. Chinese input produces Chinese scene, action, camera, sound, and music descriptions; fixed H3 field names and language tags remain protocol syntax. Preserve dialogue, lyrics, and visible scene text verbatim.","syntax":{"shot":"[Shot N] only for an actual edited shot boundary","timed_shot":"[Shot N] At MM:SS.mmm, ... for an actual later cut","timed_event":"At MM:SS.mmm, ... for a timed event inside the current shot","continuity":"Use one [Shot 1] for continuous camera/scene action; expression, dialogue, and action phases do not create new shots.","speaker":"(S1) or (S1,S2)","dialogue":"<d>[Language] exact dialogue</d>","cross_cut_dialogue":"<scenetrans>","truncated_dialogue":"<cutoff>","visible_text":"Preserve visible text verbatim inside English double quotes.","voiceover":"Use 'says in an off-screen voiceover' and state that the corresponding on-screen character's lips remain completely closed."},"camera_guidance":{"motion_types":["Zoom In","Zoom Out","Push In","Pull Out","Pan Left","Pan Right","Truck Left","Truck Right","Tilt Up","Tilt Down","Pedestal Up","Pedestal Down","Arc Shot","Tracking Shot","Static Shot","Shake Slightly","Shake Strongly","POV","Roll Clockwise","Roll Counterclockwise"],"amplitude":["with small amplitude","with large amplitude"],"speed":["at slow speed","at fast speed"],"rule":"Write camera movement as a natural action within its shot."},"format":"base","section_order":["integrated_multimodal_description","overall_soundscape","non_diegetic_music"],"keyframe_instruction":"How the reference pictures align with the target video - <Picture 1> (from [Shot N]) aligns with the S.SS-second mark of the target video.","media_labels":{"images":["<Picture 1>"]}},"ref2va":{"passthrough":"verbatim when prompt_optimization is false","structured_prompt_required":false,"optimization":{"parameter":"prompt_optimization","type":"boolean","default":false,"behavior":"When true, the service rewrites the prompt into the official H3 structure while preserving user intent, dialogue, lyrics, and visible text."},"content_review":{"always_applied":true,"stages":["original","optimized_when_present"],"rejection_status":422,"rejection_code":"prompt_review_rejected","unavailable_status":503,"unavailable_code":"prompt_review_unavailable","rewrites_prompt":false},"language_rule":"Write all prose in all six Ref2VA sections in English, regardless of the user's input language. Preserve only dialogue, lyrics, and visible scene text verbatim in their original language.","syntax":{"shot":"[Shot N] only for an actual edited shot boundary","timed_shot":"[Shot N] At MM:SS.mmm, ... for an actual later cut","timed_event":"At MM:SS.mmm, ... for a timed event inside the current shot","continuity":"Use one [Shot 1] for continuous camera/scene action; expression, dialogue, and action phases do not create new shots.","speaker":"(S1) or (S1,S2)","dialogue":"<d>[Language] exact dialogue</d>","cross_cut_dialogue":"<scenetrans>","truncated_dialogue":"<cutoff>","visible_text":"Preserve visible text verbatim inside English double quotes.","voiceover":"Use 'says in an off-screen voiceover' and state that the corresponding on-screen character's lips remain completely closed.","subject":"<Subject N>","picture":"<Picture N>","audio":"<Audio N>"},"camera_guidance":{"motion_types":["Zoom In","Zoom Out","Push In","Pull Out","Pan Left","Pan Right","Truck Left","Truck Right","Tilt Up","Tilt Down","Pedestal Up","Pedestal Down","Arc Shot","Tracking Shot","Static Shot","Shake Slightly","Shake Strongly","POV","Roll Clockwise","Roll Counterclockwise"],"amplitude":["with small amplitude","with large amplitude"],"speed":["at slow speed","at fast speed"],"rule":"Write camera movement as a natural action within its shot."},"format":"full_reference","section_order":["subject_definitions","summary","retention_analysis","detailed_description","overall_soundscape","non_diegetic_music"],"media_labels":{"images":"<Picture N> in images upload order","audios":"<Audio N> in audios upload order","numbering":"Picture, Audio, and Subject labels use independent counters."},"summary_task_types":["keyframe completion","reference generation","audio reuse","audio reference"],"retention_markers":{"visual":["fully_preserved","partially_preserved","attribute_transfer","weak_reference"],"audio":["fully_copy","partially_copy","reference","weak_reference"]},"detailed_description_budget":{"target_words_per_second":35,"maximum_words":500,"priority":"reference-role and timeline fidelity over word count"}}},"limits":{"prompt_max_chars":30000,"seconds_min":5,"seconds_max":15,"seconds_values":[5,6,7,8,9,10,11,12,13,14,15],"max_images":9,"max_videos":0,"max_audios":3,"max_media":12,"max_asset_size_bytes":52428800,"max_image_size_bytes":20971520,"max_video_size_bytes":0,"max_audio_size_bytes":52428800,"max_upload_total_size_bytes":125829120,"max_multipart_body_size_bytes":134217728},"media_preprocessing":{"images":{"automatic":true,"trigger_size_bytes":8388608,"max_long_edge":2048,"output_format":"image/jpeg"}}},"output":{"content_type":"video/mp4","fps":24,"status_endpoint":"/v1/videos/{id}","content_endpoint":"/v1/videos/{id}/content"},"billing":{"currency":"CNY","unit":"output_second","aspect_ratio_independent":true,"price_per_second_cny":{"480p":0.14,"720p":0.15}}}]}