Fitting MiniMax output into YouTube ad formats
The MiniMax reference publishes two models with deliberately different edges: four to fifteen seconds at 768P or 2K on MiniMax-H3, five to fifteen at 480P or 768P on H3-Max, where four seconds and 2K are both stated as unsupported. As of 2026-09-22.
| Format | Generations needed | Note |
|---|---|---|
| Bumper, 5-6 s | One generation | Four seconds is the floor on MiniMax-H3 |
| Non-skippable, up to 60 s | Four generations | Fifteen seconds is the ceiling on both models |
| Skippable, 30 s of paid view | Two generations | Two runs at the ceiling |
| Masthead, 16:9 up to 30 s | Two generations | 16:9 is listed, and cannot be left adaptive |
Inclusion rule. Formats from Google's own help page, matched only against limits the tool vendor publishes. Where either side does not state a value, the row says so rather than estimating. Order. Shortest format first, as on the placements page.
1How much each tool puts in writing
| Tool | Limits published |
|---|---|
| Amazon Nova Reel | 4 published |
| CogVideoX | 3 published |
| Google Veo | 4 published |
| Grok Imagine | 4 published |
| Hedra | 3 published |
| HeyGen | 4 published |
| LTX-Video | 3 published |
| MiniMax | 3 published |
| Pika | 2 published |
| PixVerse | 3 published |
| Runway | 4 published |
| SceneMixer | 3 published |
| Wan 3.0 | 4 published |
Inclusion rule. Tools with a sheet on this site, counted by how many separate output limits their own pages state. A limit mentioned only in a launch post is not counted. Order. Alphabetical by tool name.
| Tool | Bumper, 5-6 s |
|---|---|
| Amazon Nova Reel | One shot |
| CogVideoX | One generation |
| Google Veo | One generation |
| Grok Imagine | One generation |
| Hedra | One generation |
| HeyGen | One scene |
| LTX-Video | One generation |
| MiniMax | One generation |
| Pika | One generation |
| PixVerse | One generation |
| Runway | One generation |
| SceneMixer | One generation |
| Wan 3.0 | One generation |
Inclusion rule. Tools publishing output limits on their own product pages. A tool silent on this particular limit keeps its row, marked hollow. Order. Alphabetical by tool name.
| Tool | Non-skippable, up to 60 s |
|---|---|
| Amazon Nova Reel | One request of ten shots |
| CogVideoX | Ten generations |
| Google Veo | Eight generations |
| Grok Imagine | Four generations |
| Hedra | One generation |
| HeyGen | One scene |
| LTX-Video | Seven generations |
| MiniMax | Four generations |
| Pika | Six or more generations |
| PixVerse | Four generations |
| Runway | Four generations |
| SceneMixer | Two generations on Wan 3.0 |
| Wan 3.0 | Two generations |
Inclusion rule. Tools publishing output limits on their own product pages. A tool silent on this particular limit keeps its row, marked hollow. Order. Alphabetical by tool name.
| Tool | Skippable, 30 s of paid view |
|---|---|
| Amazon Nova Reel | One request of five shots |
| CogVideoX | Five generations |
| Google Veo | Four generations |
| Grok Imagine | Two generations |
| Hedra | One generation |
| HeyGen | One scene |
| LTX-Video | Four generations |
| MiniMax | Two generations |
| Pika | Two to three generations |
| PixVerse | Two generations |
| Runway | Two generations |
| SceneMixer | One generation on Wan 3.0 |
| Wan 3.0 | One generation |
Inclusion rule. Tools publishing output limits on their own product pages. A tool silent on this particular limit keeps its row, marked hollow. Order. Alphabetical by tool name.
2A reference that publishes what each model refuses
Most pages in this register list what is available. This one lists exclusions alongside: four seconds is not supported on H3-Max, 2K is not supported on H3-Max. Those sentences carry more planning information than the ranges do, because they say which of two nearly identical models a given brief has to use.
A five second bumper runs on either. A four second one runs on only one of them, and a piece that needs 2K runs on the other, so a campaign with both requirements is a campaign that has already decided it is generating twice. None of that is visible from a price per second.
3The ratio is required, which is rarer than it sounds
For text-to-video the reference states that the ratio is required and cannot be adaptive, and gives six values from 21:9 down to 9:16. Required is the useful word: a request that omits the shape does not quietly receive widescreen, it does not run. On a site whose recurring finding is that shape gets decided by a default nobody read, that is a design choice worth pointing at.
For image-to-video the same page says the ratio follows the input image and is always adaptive, which moves the decision to whoever prepared the reference frame. The shape is still determined before generation, but by a file rather than by a parameter, and that file is usually chosen for its content rather than its dimensions.
4Sources
Output limits read from the video API reference at platform.minimax.io, checked 2026-09-22. Format requirements are from Google Ads Help, and the matching rule is on method. Next along the register: Pika, PixVerse. The whole set of 15 is on published limits, and the sheets for each are indexed at tools.
- Durations4-15 seconds on MiniMax-H3, and 5-15 on H3-Max where 4 seconds is not supportedstated per model
- Resolutions768P and 2K on MiniMax-H3; 480P and 768P on H3-Max, where 2K is not supportedstated per model
- Aspect ratios; for image-to-video the ratio follows the input image21:9, 16:9, 4:3, 1:1, 3:4 and 9:16, required and never adaptive for text-to-video