> ## Documentation Index > Fetch the complete documentation index at: https://powower.mintlify.site/llms.txt > Use this file to discover all available pages before exploring further. # First Frame to Video Generation > Video generation V2 API. Input via multimodal content array (text / image / video / audio). Supports text-to-video, image-to-video (first/last frame), and multimodal reference-to-video with 2K output. ## OpenAPI ````yaml api-reference/en/zmodelVideo/minimax/minimax-h3/first-frame-to-video.json POST /minimax/v2/video_generation openapi: 3.0.1 info: title: MiniMax API description: Create a video generation task. version: 1.0.0 servers: - url: https://baze-api.powerbuyin.top security: [] paths: /minimax/v2/video_generation: post: summary: Image-to-video (t2va) description: >- Video generation V2 API. Input via multimodal content array (text / image / video / audio). Supports text-to-video, image-to-video (first/last frame), and multimodal reference-to-video with 2K output. operationId: videoGeneration parameters: - name: Content-Type in: header required: true description: Media type of the request body. Set to `application/json`. schema: type: string enum: - application/json default: application/json requestBody: required: true description: '' content: application/json: schema: $ref: '#/components/schemas/VideoGenerationReq' examples: Image-to-video (t2va): summary: Image-to-video (t2va) value: model: MiniMax-H3 content: - type: text text: >- Pull focus to the people in the background and add more steam to the ramen bowl. - type: image_url image_url: url: >- https://cdn.hailuoai.com/prod/hailuo_demo/testsets/H3_AA_I2VA/gallery/sr_v17_variants_seed42_43_20260724/inputs/4a3a90bf9100_KDmcbkhzYo5sjjxr9FqcVmWVnzb.png role: first_frame resolution: 2K duration: 5 ratio: adaptive responses: '200': description: >- Returns task_id upon successful creation. Use this task_id to call the query task API to retrieve task status and results. Query task success response example ```json { "task": { "id": "424010985738629", "model": "MiniMax-H3", "status": "succeeded", "created_at": 1785125529, "updated_at": 1785125946, "content": { "url": "https://your-cdn.example.com/h3-generated-2k-output.mp4" }, "resolution": "2K", "duration": 5, "usage": { "total_seconds": 5, "input_seconds": 0, "output_seconds": 5, "input_image_count": 0 }, "ratio": "16:9", "task_type": "generation", "modality": "video" } } content: application/json: schema: $ref: '#/components/schemas/VideoGenerationResp' example: output: task_id: '424010985738629' security: - bearerAuth: [] components: schemas: VideoGenerationReq: type: object required: - model - content - resolution - duration properties: model: type: string description: 'Model name. Currently available: `MiniMax-H3`.' enum: - MiniMax-H3 content: type: array description: > Multimodal input content array describing the information used to generate the video. Each element is distinguished by type (text / image_url / video_url / audio_url) and can be annotated with a role. Each request must contain a non-empty text item (prompt is required); omission will return a parameter error. Supported input combinations (corresponding to different generation scenarios): Image-to-video (first frame): text + 1 image_url (role=first_frame or omitted). Image-to-video (last frame): text + 1 image_url (role=last_frame). Image-to-video (first-last frame): text + 2 image_url (roles first_frame and last_frame respectively). Image-to-video and multimodal reference-to-video are mutually exclusive: if any of reference_image / reference_video / reference_audio roles appear in content, first_frame / last_frame must not appear (and vice versa); they cannot be mixed. Input media limits (total request body size ≤ 64 MB; use public URLs for large files, do not use Base64) Image image_url: | Item | Limit | | :--- | :--- | | Format | JPG, JPEG, PNG, WEBP, HEIC, HEIF | | Single file size | ≤ 30 MB | | Width/height range | [256, 5760] px | | Aspect ratio (w/h) | [0.4, 2.5] | | Count | first frame ≤ 1, last frame ≤ 1, reference images ≤ 9 | items: type: object required: - type properties: type: type: string description: Type of the input content. enum: - text text: type: string description: >- Text prompt. All scenarios require a non-empty text describing the expected video. Single text max 7000 characters. image_url: type: object description: >- Image object when type=image_url (format / size / dimension / count limits see content description above). required: - url properties: url: type: string description: >- Image address. Supports: public URL; mm_file://{file_id} (reference platform existing files, such as uploaded or historical output file_id); data:image/;base64, data URI ( in lowercase). role: type: string description: >- The position or purpose of the content, conditionally required: - first_frame: First frame image (image-to-video; if only one image is provided and role is not set, it defaults to first_frame). - last_frame: Last frame image (image-to-video first-last frame; must be paired with first_frame). - reference_image: Reference image (multimodal reference-to-video). - reference_video: Reference video (multimodal reference-to-video). - reference_audio: Reference audio (multimodal reference-to-video). enum: - first_frame - last_frame - reference_image - reference_video - reference_audio resolution: type: string description: 'Video resolution. Currently available: 768P, 2K.' enum: - 768P - 2K duration: type: integer description: 'Video duration in seconds. Required, integer. Available: 4~15.' enum: - 4 - 5 - 6 - 7 - 8 - 9 - 10 - 11 - 12 - 13 - 14 - 15 ratio: type: string description: >- Output video aspect ratio, default adaptive (automatically selects the most suitable aspect ratio based on input; the actual ratio can be retrieved from the ratio field in the query API). Text-to-video (t2va, content contains only text): ratio is required and cannot be adaptive; available values: 21:9, 16:9, 4:3, 1:1, 3:4, 9:16. enum: - '21:9' - '16:9' - '4:3' - '1:1' - '3:4' - '9:16' callback_url: type: string description: >- Callback notification URL for task status changes. After configuration, the MiniMax server will first send a verification request containing a challenge field (must return the challenge value unchanged within 3 seconds to complete verification). Once verified, whenever the task status changes, a POST push will be sent to this URL, with the push body structure matching the query task API response. Callback status values: queued, running, succeeded, failed, cancelled. aigc_watermark: type: boolean description: >- Whether to add an AIGC watermark to the generated video. Default false. VideoGenerationResp: type: object properties: task_id: type: string description: Task ID, used to query task status and results. securitySchemes: bearerAuth: type: http scheme: bearer bearerFormat: JWT description: 'Pass `Authorization: Bearer ` in the request header.' ````