MiniMax H3 Max Referensi ke Video API: Gambar-ke-Video

Create video dari gambar, video, dan audio referensis dengan MiniMax H3 Max on Muapi. Up ke 1080p, pay per generation. MiniMax H3 Max Referensi ke Video menghasilkans a video clip guided by a prompt plus up ke 9 referensi gambars, 3 videos, dan 3 audio clips, dengan no more than 12 referensi files combined. Reference assets by name in Anda prompt (for example "Image 1" or "Video 1") ke tell model how each one sebaiknya influence result — a character's appearance, a background style, a gerakan pattern, or a soundtrack. Output resolusi runs up ke 1080p across durasis of 5-15 seconds, dengan an adaptive aspect ratio option that follows referensis automatically. It follows Muapi's standard asynchronous request/poll/webhook pattern. For a single starting gambar without multi-referensi blending, see MiniMax H3 Max Image to Video; untuk a pure text prompt dengan no referensis, see MiniMax H3 Max Text to Video.

Kontrol model interaktif

Create video dari gambar, video, dan audio referensis dengan MiniMax H3 Max on Muapi. Up ke 1080p, pay per generation.

📝

Ikhtisar

Tentang model ini

MiniMax H3 Max Referensi ke Video menghasilkans a video clip guided by a prompt plus up ke 9 referensi gambars, 3 videos, dan 3 audio clips, dengan no more than 12 referensi files combined. Reference assets by name in Anda prompt (for example "Image 1" or "Video 1") ke tell model how each one sebaiknya influence result — a character's appearance, a background style, a gerakan pattern, or a soundtrack. Output resolusi runs up ke 1080p across durasis of 5-15 seconds, dengan an adaptive aspect ratio option that follows referensis automatically. It follows Muapi's standard asynchronous request/poll/webhook pattern. For a single starting gambar without multi-referensi blending, see MiniMax H3 Max Image to Video; untuk a pure text prompt dengan no referensis, see MiniMax H3 Max Text to Video.

1Character consistency: Gunakan multiple referensi gambars of same character ke keep their appearance consistent across a new scene.
2Style transfer: Blend a referensi video's gerakan or camera style dengan new subject matter.
3Music-driven edits: Layer a referensi audio track alongside visual referensis ke guide pacing dan mood.
4Product reveals: Combine product referensi gambars dengan a gerakan referensi video untuk a polished reveal shot.
💰

Harga & Nilai

Analisis biaya

muapiappFrom $0.065/sec (480p) up to $0.208/sec (1080p), plus reference item fees

Per-second output biaya by resolusi; first 4 referensi items adalah included, each additional referensi gambar/video/audio item adds a small flat fee. Pay-per-generation, no subscription.

ReplicateTidak tersedia

MiniMax H3 Max adalah not listed on Replicate.

** Harga pesaing diperkirakan berdasarkan arsitektur model dan tingkat penggunaan yang serupa.

⚙️

Detail Teknis

Skema konfigurasi

cepatstring

Describe the video scene and how the references should be used (e.g. 'Image 1', 'Video 1').

Nilai BawaanUse the references to create a cinematic product reveal with a slow camera orbit.
Gambar referensiarray

Reference image URLs (up to 9). Provide at least one image or video reference. Max 12 combined reference files.

Nilai Bawaanundefined
Video Referensiarray

Reference video URLs (up to 3; 2-15s each, 15s combined max). Provide at least one image or video reference.

Nilai Bawaanundefined
Audio Referensiarray

Optional reference audio URLs (up to 3; 2-15s each, 15s combined max). Audio cannot be used alone.

Nilai Bawaanundefined
rasio aspekEnum (opsi 7)

-

Nilai Bawaanadaptive
ResolusiEnum (opsi 3)

-

Nilai Bawaan768p
Durasi (detik)Enum (opsi 11)

durasi keluaran () .

Nilai Bawaan5
Prompt Expansion ModeEnum (opsi 3)

Prompt rewriting effort. 'quality' can add up to ~30s of processing.

Nilai Bawaanbalanced
📖

Panduan Implementasi

Dokumentasi pengembang

Cara Menggunakan MiniMax H3 Max Referensi ke Video

  1. Add referensis: Provide reference_images (up ke 9) and/or reference_videos (up ke 3) URLs; reference_audios (up ke 3) adalah optional dan cannot be used alone. The combined total across all three fields cannot exceed 12 files, dan pada least one gambar or video adalah required.

  2. Write prompt: Describe scene dan explicitly referensi Anda assets by name, e.g. "Gunakan Image 1's character walking through scene dari Video 1."

  3. Pick resolusi dan durasi: Set resolution ke 480p, 768p, or 1080p, dan duration ke any value dari 5 ke 15 seconds.

  4. Set aspect ratio: Leave aspect_ratio as adaptive ke follow referensis automatically, or choose a fixed ratio.

  5. Submit dan retrieve: POST request, then poll /predictions/{request_id}/result or supply a webhook_url ke receive finished video URL.

curl -X POST https://api.muapi.ai/api/v1/minimax-h3-max-reference-to-video \
  -H "x-api-key: YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "prompt": "Use the references to create a cinematic product reveal with a slow camera orbit.",
    "reference_images": ["https://example.com/product.jpg"],
    "resolution": "1080p",
    "duration": 6
  }'
❓

Pertanyaan Umum

Sering ditanyakan

Apa yang dihasilkan MiniMax H3 Max Referensi ke Video menghasilkan?

It menghasilkans a video guided by a prompt plus one or more referensi gambars, videos, dan optional audio, blending their appearance, gerakan, or sound into new clip.

Bagaimana many referensi files dapat I use?

Up ke 12 combined referensi gambars dan videos, plus optional referensi audio tracks up ke 15 seconds combined.

Dapatkah saya use referensi audio alone?

Tidak, referensi audio must be paired dengan pada least one referensi gambar or video.

Resolusi dan durasi apa yang didukung?

Resolusi dapat berupa 480p, 768p, atau 1080p, dan durasi dapat bernilai antara 5 hingga 15 detik.

Bagaimana does adaptive aspect ratio work?

When `aspect_ratio` adalah left as `adaptive`, output follows aspect ratio implied by Anda referensi assets instead of a fixed value.

Bagaimana cara mengambil video yang sudah selesai?

Lakukan polling menggunakan ID permintaan yang dikembalikan oleh POST awal di `/predictions/{request_id}/result`, atau berikan `webhook_url` agar mendapat notifikasi saat video siap.