MiniMax H3 Max
The premium tier of the native-audio video model
- frameImages
- video
MiniMax H3 Max is the step-up tier of the model that brought native stereo audio to video generation, and this node runs it over Runware's cloud. If you've read the base H3 page, you know the setup: same pack, same shape, and here a leaner input list with a smaller default size - 480p rather than 768p - which is your first hint that Max is a quality-first, cost-aware offering aimed at renders you're keeping, not drafts you're burning.
Quick recap for context, because it decides how you use this. H3 is MiniMax's 33B omni-modal model - text, image, video, and audio treated as one input, output video with synchronized native sound, up to 15 seconds - and its open-weights release took #1 in video editing with audio on Artificial Analysis while carrying a community license that geofences the US, EU, UK, and South Korea. The base H3 page makes the argument for the hosted route, and it holds here unchanged: Max through Runware means no territory-restricted 42GB download and no VRAM ceiling, just a metered API call from inside your workflow.
What differs on Max is the control surface. The creative essentials stay: required positivePrompt, frameImages for frame conditioning, seed, numberResults for variations, and duration (5–15 seconds, default 5). Notably, Max's widget list is shorter than the base node's - no referenceVideos socket and no advanced_json escape hatch, so if you came looking for video-to-video transfer or raw JSON injection, this tier isn't the one that carries them; the base H3 node is. You get size (presets starting at 480p), the safety group with fast/full modes, outputFormat (MP4/WEBM/MOV), outputQuality, and TTL settings. As with the base, there's no audio toggle - native stereo sound is the model's baseline. Output: VIDEO.
Install
The same one-time Runware pack setup: ComfyUI Manager → search Runware → install → restart, or clone https://github.com/Runware/ComfyUI-Runware into custom_nodes and pip install -r requirements.txt. Then a Runware API key from the dashboard (Settings → Runware API key, RUNWARE_API_KEY, or runware auth login).
Picking between the H3 tiers
Use Max for final renders where the base tier's output came up short and the subject justifies the premium rate; keep base H3 for iteration, since it carries the video-to-video socket and the advanced_json escape hatch that Max drops. Every cloud-video caveat applies with force at the premium tier - watch the per-run cost on the node's title bar, and remember your prompt and frames transit Runware and MiniMax's moderation. There's no local fallback for most Western users anyway, which is precisely why the node exists.
Inputs (13)
| Name | Type | Default | Description |
|---|---|---|---|
| positivePrompt | STRING | Text prompt describing elements to include in the generated output. | |
| frameImagesopt | IMAGE | — | |
| sizeopt | COMBO | 480p (~16:9) | Output size. Only one of resolution or dimensions is sent. |
| seedopt | INT | 00–4294967295 | Random seed for reproducible generation. When not provided, a random seed is generated in the unsigned 32-bit range. |
| durationopt | INT | 55–15 | Length of the generated video in seconds. The total number of frames produced is determined by duration multiplied by the model's frame rate (fps). |
| numberResultsopt | INT | 11–4 | Number of results to generate. Each result uses a different seed, producing variations of the same parameters. |
| safetyopt | BOOLEAN | false | Enable to set safety. Off uses the model's default. |
| safety.checkContentopt | BOOLEAN | false | Enable or disable content safety checking. Increases total generation time. |
| safety.modeopt | COMBO | fast | Safety checking mode for video generation. |
| ttlopt | BOOLEAN | false | Enable to set ttl. Off uses the model's default. |
| ttl_valueopt | INT | 60 | Time-to-live (TTL) in seconds for generated content. Only applies when `outputType` is `URL`. |
| outputFormatopt | COMBO | MP4 | File format for the generated video. |
| outputQualityopt | INT | 9520–99 | Compression quality of the output. Higher values preserve quality but increase file size. |
Outputs (1)
| Name | Type | Description |
|---|---|---|
| video | VIDEO | — |