> ## Documentation Index > Fetch the complete documentation index at: https://oxenai-eric-fix-queue-reference-claims.mintlify.site/llms.txt > Use this file to discover all available pages before exploring further. # MiniMax H3 Max: Reference to Video > Reference-guided video with lip-synced audio Run this model interactively, tune parameters, and compare outputs. **Model ID:** `minimax-h3-max-reference-to-video` MiniMax H3 Max (Reference to Video) is a post-trained variant of MiniMax H3, tuned by fal for stronger prompt adherence and better aesthetics. It generates video from a text prompt guided by multimodal references: up to 9 subject or style images, 3 motion video clips, and 3 audio clips. It generates 5 to 15 second clips at 480P or 768P resolution with audio, including lip-synced dialogue, keeping subjects consistent with their reference images while following referenced motion and voices. References may total at most 12 files; video and audio clips run 2 to 15 seconds each with at most 15 seconds combined per type, and audio cannot be the only reference. Prompt expansion modes trade latency for prompt fidelity. ## Example request Use the [Workbench](https://www.oxen.ai/ai/workbench?model=minimax-h3-max-reference-to-video) as a request builder: configure parameters for this model in the UI, then open the **API** tab to copy the exact cURL or Python call. This blocks until the video is ready (typically 5-15 minutes). Prefer **Async** or **Async with SSE** for anything beyond quick experimentation. See the [video generation reference](/inference-api/reference/video_generation) for more details. ```bash cURL theme={null} curl -X POST https://hub.oxen.ai/api/ai/videos/generate \ -H "Content-Type: application/json" \ -H "Authorization: Bearer $OXEN_API_KEY" \ -d '{ "model": "minimax-h3-max-reference-to-video", "prompt": "" }' ``` ```python Python theme={null} import os import requests response = requests.post( "https://hub.oxen.ai/api/ai/videos/generate", headers={ "Content-Type": "application/json", "Authorization": f"Bearer {os.environ['OXEN_API_KEY']}", }, json={ "model": "minimax-h3-max-reference-to-video", "prompt": "" }, ) response.raise_for_status() print(response.json()) ``` ```bash cURL theme={null} curl -X POST https://hub.oxen.ai/api/ai/videos/generate \ -H "Content-Type: application/json" \ -H "Authorization: Bearer $OXEN_API_KEY" \ -d '{ "model": "minimax-h3-max-reference-to-video", "prompt": "", "input_images": [ "https://hub.oxen.ai/api/repos/elau/assets/file/main/bloxy/bloxy_cropped_512x512.png" ], "input_videos": [ "https://hub.oxen.ai/api/repos/ox/Oxen-AI-Assets/file/main/images/winter_summer_ox.mp4" ], "input_audios": [ "https://example.com/audio.mp3" ] }' ``` ```python Python theme={null} import os import requests response = requests.post( "https://hub.oxen.ai/api/ai/videos/generate", headers={ "Content-Type": "application/json", "Authorization": f"Bearer {os.environ['OXEN_API_KEY']}", }, json={ "model": "minimax-h3-max-reference-to-video", "prompt": "", "input_images": [ "https://hub.oxen.ai/api/repos/elau/assets/file/main/bloxy/bloxy_cropped_512x512.png" ], "input_videos": [ "https://hub.oxen.ai/api/repos/ox/Oxen-AI-Assets/file/main/images/winter_summer_ox.mp4" ], "input_audios": [ "https://example.com/audio.mp3" ] }, ) response.raise_for_status() print(response.json()) ``` ```bash cURL theme={null} curl -X POST https://hub.oxen.ai/api/ai/videos/generate \ -H "Content-Type: application/json" \ -H "Authorization: Bearer $OXEN_API_KEY" \ -d '{ "model": "minimax-h3-max-reference-to-video", "prompt": "", "input_images": [ "https://hub.oxen.ai/api/repos/elau/assets/file/main/bloxy/bloxy_cropped_512x512.png" ], "input_videos": [ "https://hub.oxen.ai/api/repos/ox/Oxen-AI-Assets/file/main/images/winter_summer_ox.mp4" ], "input_audios": [ "https://example.com/audio.mp3" ], "aspect_ratio": "adaptive", "resolution": "768P", "duration": 5, "prompt_expansion_mode": "balanced", "enable_safety_checker": true, "sync_mode": false }' ``` ```python Python theme={null} import os import requests response = requests.post( "https://hub.oxen.ai/api/ai/videos/generate", headers={ "Content-Type": "application/json", "Authorization": f"Bearer {os.environ['OXEN_API_KEY']}", }, json={ "model": "minimax-h3-max-reference-to-video", "prompt": "", "input_images": [ "https://hub.oxen.ai/api/repos/elau/assets/file/main/bloxy/bloxy_cropped_512x512.png" ], "input_videos": [ "https://hub.oxen.ai/api/repos/ox/Oxen-AI-Assets/file/main/images/winter_summer_ox.mp4" ], "input_audios": [ "https://example.com/audio.mp3" ], "aspect_ratio": "adaptive", "resolution": "768P", "duration": 5, "prompt_expansion_mode": "balanced", "enable_safety_checker": true, "sync_mode": false }, ) response.raise_for_status() print(response.json()) ``` See the [async queue reference](/inference-api/reference/async_queue) for more details. ```bash cURL theme={null} # Enqueue, capture the generation id. GEN_ID=$(curl -s -X POST https://hub.oxen.ai/api/ai/queue \ -H "Content-Type: application/json" \ -H "Authorization: Bearer $OXEN_API_KEY" \ -d '{ "model": "minimax-h3-max-reference-to-video", "prompt": "" }' | jq -r '.generations[0].generation_id') # Poll until the generation reaches a terminal status. while true; do STATUS=$(curl -s -H "Authorization: Bearer $OXEN_API_KEY" \ "https://hub.oxen.ai/api/ai/queue/$GEN_ID" | jq -r '.status') echo "Status: $STATUS" case $STATUS in succeeded|failed|cancelled) break;; esac sleep 5 done # Print the result. curl -s -H "Authorization: Bearer $OXEN_API_KEY" \ "https://hub.oxen.ai/api/ai/queue/$GEN_ID" | jq . ``` ```python Python theme={null} import os import time import requests HEADERS = { "Content-Type": "application/json", "Authorization": f"Bearer {os.environ['OXEN_API_KEY']}", } enqueue = requests.post( "https://hub.oxen.ai/api/ai/queue", headers=HEADERS, json={ "model": "minimax-h3-max-reference-to-video", "prompt": "" }, ) enqueue.raise_for_status() generation_id = enqueue.json()["generations"][0]["generation_id"] while True: data = requests.get( f"https://hub.oxen.ai/api/ai/queue/{generation_id}", headers=HEADERS, ).json() if data["status"] in {"succeeded", "failed", "cancelled"}: break time.sleep(5) if data["status"] == "succeeded": print(f"Result: {data['result_url']}") else: print(f"Generation {data['status']}: {data.get('error_message')}") ``` ```bash cURL theme={null} # Enqueue, capture the generation id. GEN_ID=$(curl -s -X POST https://hub.oxen.ai/api/ai/queue \ -H "Content-Type: application/json" \ -H "Authorization: Bearer $OXEN_API_KEY" \ -d '{ "model": "minimax-h3-max-reference-to-video", "prompt": "", "input_images": [ "https://hub.oxen.ai/api/repos/elau/assets/file/main/bloxy/bloxy_cropped_512x512.png" ], "input_videos": [ "https://hub.oxen.ai/api/repos/ox/Oxen-AI-Assets/file/main/images/winter_summer_ox.mp4" ], "input_audios": [ "https://example.com/audio.mp3" ] }' | jq -r '.generations[0].generation_id') # Poll until the generation reaches a terminal status. while true; do STATUS=$(curl -s -H "Authorization: Bearer $OXEN_API_KEY" \ "https://hub.oxen.ai/api/ai/queue/$GEN_ID" | jq -r '.status') echo "Status: $STATUS" case $STATUS in succeeded|failed|cancelled) break;; esac sleep 5 done # Print the result. curl -s -H "Authorization: Bearer $OXEN_API_KEY" \ "https://hub.oxen.ai/api/ai/queue/$GEN_ID" | jq . ``` ```python Python theme={null} import os import time import requests HEADERS = { "Content-Type": "application/json", "Authorization": f"Bearer {os.environ['OXEN_API_KEY']}", } enqueue = requests.post( "https://hub.oxen.ai/api/ai/queue", headers=HEADERS, json={ "model": "minimax-h3-max-reference-to-video", "prompt": "", "input_images": [ "https://hub.oxen.ai/api/repos/elau/assets/file/main/bloxy/bloxy_cropped_512x512.png" ], "input_videos": [ "https://hub.oxen.ai/api/repos/ox/Oxen-AI-Assets/file/main/images/winter_summer_ox.mp4" ], "input_audios": [ "https://example.com/audio.mp3" ] }, ) enqueue.raise_for_status() generation_id = enqueue.json()["generations"][0]["generation_id"] while True: data = requests.get( f"https://hub.oxen.ai/api/ai/queue/{generation_id}", headers=HEADERS, ).json() if data["status"] in {"succeeded", "failed", "cancelled"}: break time.sleep(5) if data["status"] == "succeeded": print(f"Result: {data['result_url']}") else: print(f"Generation {data['status']}: {data.get('error_message')}") ``` ```bash cURL theme={null} # Enqueue, capture the generation id. GEN_ID=$(curl -s -X POST https://hub.oxen.ai/api/ai/queue \ -H "Content-Type: application/json" \ -H "Authorization: Bearer $OXEN_API_KEY" \ -d '{ "model": "minimax-h3-max-reference-to-video", "prompt": "", "input_images": [ "https://hub.oxen.ai/api/repos/elau/assets/file/main/bloxy/bloxy_cropped_512x512.png" ], "input_videos": [ "https://hub.oxen.ai/api/repos/ox/Oxen-AI-Assets/file/main/images/winter_summer_ox.mp4" ], "input_audios": [ "https://example.com/audio.mp3" ], "aspect_ratio": "adaptive", "resolution": "768P", "duration": 5, "prompt_expansion_mode": "balanced", "enable_safety_checker": true, "sync_mode": false }' | jq -r '.generations[0].generation_id') # Poll until the generation reaches a terminal status. while true; do STATUS=$(curl -s -H "Authorization: Bearer $OXEN_API_KEY" \ "https://hub.oxen.ai/api/ai/queue/$GEN_ID" | jq -r '.status') echo "Status: $STATUS" case $STATUS in succeeded|failed|cancelled) break;; esac sleep 5 done # Print the result. curl -s -H "Authorization: Bearer $OXEN_API_KEY" \ "https://hub.oxen.ai/api/ai/queue/$GEN_ID" | jq . ``` ```python Python theme={null} import os import time import requests HEADERS = { "Content-Type": "application/json", "Authorization": f"Bearer {os.environ['OXEN_API_KEY']}", } enqueue = requests.post( "https://hub.oxen.ai/api/ai/queue", headers=HEADERS, json={ "model": "minimax-h3-max-reference-to-video", "prompt": "", "input_images": [ "https://hub.oxen.ai/api/repos/elau/assets/file/main/bloxy/bloxy_cropped_512x512.png" ], "input_videos": [ "https://hub.oxen.ai/api/repos/ox/Oxen-AI-Assets/file/main/images/winter_summer_ox.mp4" ], "input_audios": [ "https://example.com/audio.mp3" ], "aspect_ratio": "adaptive", "resolution": "768P", "duration": 5, "prompt_expansion_mode": "balanced", "enable_safety_checker": true, "sync_mode": false }, ) enqueue.raise_for_status() generation_id = enqueue.json()["generations"][0]["generation_id"] while True: data = requests.get( f"https://hub.oxen.ai/api/ai/queue/{generation_id}", headers=HEADERS, ).json() if data["status"] in {"succeeded", "failed", "cancelled"}: break time.sleep(5) if data["status"] == "succeeded": print(f"Result: {data['result_url']}") else: print(f"Generation {data['status']}: {data.get('error_message')}") ``` See the [async queue reference](/inference-api/reference/async_queue) for more details. ```bash cURL theme={null} # Enqueue, capture the generation id. GEN_ID=$(curl -s -X POST https://hub.oxen.ai/api/ai/queue \ -H "Content-Type: application/json" \ -H "Authorization: Bearer $OXEN_API_KEY" \ -d '{ "model": "minimax-h3-max-reference-to-video", "prompt": "" }' | jq -r '.generations[0].generation_id') # Stream the SSE channel, grab the data line that follows a # media_generation_completed event for our id, and pretty-print it. curl -sN -H "Authorization: Bearer $OXEN_API_KEY" https://hub.oxen.ai/api/events \ | awk -v id="$GEN_ID" ' /^event: media_generation_completed$/ { expect=1; next } /^data: / && expect { payload = substr($0, 7) if (index(payload, "\"generation_id\":\"" id "\"")) { print payload; exit } expect = 0 } ' | jq . ``` ```python Python theme={null} import json import os import requests API_KEY = os.environ["OXEN_API_KEY"] AUTH = {"Authorization": f"Bearer {API_KEY}"} enqueue = requests.post( "https://hub.oxen.ai/api/ai/queue", headers={**AUTH, "Content-Type": "application/json"}, json={ "model": "minimax-h3-max-reference-to-video", "prompt": "" }, ) enqueue.raise_for_status() generation_id = enqueue.json()["generations"][0]["generation_id"] with requests.get( "https://hub.oxen.ai/api/events", headers=AUTH, stream=True, ) as stream: event_name = None for line in stream.iter_lines(decode_unicode=True): if line.startswith("event: "): event_name = line.removeprefix("event: ") elif line.startswith("data: ") and event_name == "media_generation_completed": payload = json.loads(line.removeprefix("data: ")) if payload.get("generation_id") == generation_id: print(payload) break ``` ```bash cURL theme={null} # Enqueue, capture the generation id. GEN_ID=$(curl -s -X POST https://hub.oxen.ai/api/ai/queue \ -H "Content-Type: application/json" \ -H "Authorization: Bearer $OXEN_API_KEY" \ -d '{ "model": "minimax-h3-max-reference-to-video", "prompt": "", "input_images": [ "https://hub.oxen.ai/api/repos/elau/assets/file/main/bloxy/bloxy_cropped_512x512.png" ], "input_videos": [ "https://hub.oxen.ai/api/repos/ox/Oxen-AI-Assets/file/main/images/winter_summer_ox.mp4" ], "input_audios": [ "https://example.com/audio.mp3" ] }' | jq -r '.generations[0].generation_id') # Stream the SSE channel, grab the data line that follows a # media_generation_completed event for our id, and pretty-print it. curl -sN -H "Authorization: Bearer $OXEN_API_KEY" https://hub.oxen.ai/api/events \ | awk -v id="$GEN_ID" ' /^event: media_generation_completed$/ { expect=1; next } /^data: / && expect { payload = substr($0, 7) if (index(payload, "\"generation_id\":\"" id "\"")) { print payload; exit } expect = 0 } ' | jq . ``` ```python Python theme={null} import json import os import requests API_KEY = os.environ["OXEN_API_KEY"] AUTH = {"Authorization": f"Bearer {API_KEY}"} enqueue = requests.post( "https://hub.oxen.ai/api/ai/queue", headers={**AUTH, "Content-Type": "application/json"}, json={ "model": "minimax-h3-max-reference-to-video", "prompt": "", "input_images": [ "https://hub.oxen.ai/api/repos/elau/assets/file/main/bloxy/bloxy_cropped_512x512.png" ], "input_videos": [ "https://hub.oxen.ai/api/repos/ox/Oxen-AI-Assets/file/main/images/winter_summer_ox.mp4" ], "input_audios": [ "https://example.com/audio.mp3" ] }, ) enqueue.raise_for_status() generation_id = enqueue.json()["generations"][0]["generation_id"] with requests.get( "https://hub.oxen.ai/api/events", headers=AUTH, stream=True, ) as stream: event_name = None for line in stream.iter_lines(decode_unicode=True): if line.startswith("event: "): event_name = line.removeprefix("event: ") elif line.startswith("data: ") and event_name == "media_generation_completed": payload = json.loads(line.removeprefix("data: ")) if payload.get("generation_id") == generation_id: print(payload) break ``` ```bash cURL theme={null} # Enqueue, capture the generation id. GEN_ID=$(curl -s -X POST https://hub.oxen.ai/api/ai/queue \ -H "Content-Type: application/json" \ -H "Authorization: Bearer $OXEN_API_KEY" \ -d '{ "model": "minimax-h3-max-reference-to-video", "prompt": "", "input_images": [ "https://hub.oxen.ai/api/repos/elau/assets/file/main/bloxy/bloxy_cropped_512x512.png" ], "input_videos": [ "https://hub.oxen.ai/api/repos/ox/Oxen-AI-Assets/file/main/images/winter_summer_ox.mp4" ], "input_audios": [ "https://example.com/audio.mp3" ], "aspect_ratio": "adaptive", "resolution": "768P", "duration": 5, "prompt_expansion_mode": "balanced", "enable_safety_checker": true, "sync_mode": false }' | jq -r '.generations[0].generation_id') # Stream the SSE channel, grab the data line that follows a # media_generation_completed event for our id, and pretty-print it. curl -sN -H "Authorization: Bearer $OXEN_API_KEY" https://hub.oxen.ai/api/events \ | awk -v id="$GEN_ID" ' /^event: media_generation_completed$/ { expect=1; next } /^data: / && expect { payload = substr($0, 7) if (index(payload, "\"generation_id\":\"" id "\"")) { print payload; exit } expect = 0 } ' | jq . ``` ```python Python theme={null} import json import os import requests API_KEY = os.environ["OXEN_API_KEY"] AUTH = {"Authorization": f"Bearer {API_KEY}"} enqueue = requests.post( "https://hub.oxen.ai/api/ai/queue", headers={**AUTH, "Content-Type": "application/json"}, json={ "model": "minimax-h3-max-reference-to-video", "prompt": "", "input_images": [ "https://hub.oxen.ai/api/repos/elau/assets/file/main/bloxy/bloxy_cropped_512x512.png" ], "input_videos": [ "https://hub.oxen.ai/api/repos/ox/Oxen-AI-Assets/file/main/images/winter_summer_ox.mp4" ], "input_audios": [ "https://example.com/audio.mp3" ], "aspect_ratio": "adaptive", "resolution": "768P", "duration": 5, "prompt_expansion_mode": "balanced", "enable_safety_checker": true, "sync_mode": false }, ) enqueue.raise_for_status() generation_id = enqueue.json()["generations"][0]["generation_id"] with requests.get( "https://hub.oxen.ai/api/events", headers=AUTH, stream=True, ) as stream: event_name = None for line in stream.iter_lines(decode_unicode=True): if line.startswith("event: "): event_name = line.removeprefix("event: ") elif line.startswith("data: ") and event_name == "media_generation_completed": payload = json.loads(line.removeprefix("data: ")) if payload.get("generation_id") == generation_id: print(payload) break ``` ## Fetch model details The [models endpoint](/inference-api/reference/models/overview) returns the full model object, including its `json_request_schema`. ```bash theme={null} curl -H "Authorization: Bearer $OXEN_API_KEY" https://hub.oxen.ai/api/ai/models/minimax-h3-max-reference-to-video ``` ## Request parameters ### Required parameters | Field | Type | Default | Description | | -------- | -------- | ------- | -------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- | | `prompt` | `string` | — | Text prompt for video generation. Use @Image1, @Image2, ... to reference the reference images in order, @Video1, ... for the reference videos, and @Audio1, ... for the reference audio; each type is numbered separately. | ### Optional parameters | Field | Type | Default | Description | | ----------------------- | --------------- | ------------ | ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- | | `input_images` | `array` | — | Reference images for subjects or style, up to 9. Order maps to @Image1, @Image2, etc. in the prompt. Reference images, videos, and audio must total at most 12 files. | | `input_videos` | `array` | — | Reference video clips for motion or style, up to 3 clips of 2 to 15 seconds (15 seconds combined). Order maps to @Video1, @Video2, etc. in the prompt. Reference images, videos, and audio must total at most 12 files. | | `input_audios` | `array` | — | Reference audio for voices or sound, up to 3 clips of 2 to 15 seconds (15 seconds combined). Order maps to @Audio1, @Audio2, etc. in the prompt. Cannot be the only reference input; provide at least one reference image or video with it. | | `aspect_ratio` | `string` | `"adaptive"` | The aspect ratio of the generated video. One of: adaptive, 21:9, 16:9, 4:3, 1:1, 3:4, 9:16. | | `resolution` | `string` | `"768P"` | The native generation resolution of the video. One of: 480P, 768P. | | `duration` | `integer` | `5` | The duration of the video in seconds. Range: 5 – 15. | | `prompt_expansion_mode` | `string` | `"balanced"` | How much effort to spend rewriting the prompt before generation. 'balanced' returns in about a second. 'quality' spends up to \~30s on a richer prompt. One of: balanced, quality. | | `enable_safety_checker` | `boolean` | `true` | If set to true, the safety checker will be enabled. | | `sync_mode` | `boolean` | `false` | Return the generated video as base64 instead of a CDN URL. | | `seed` | `integer` | — | Random seed. A random seed is selected when omitted. |