Understand the local workflow
JollyAI's reviewed H3 integration uses a resolution selector with a megapixel budget and dimensions aligned to model requirements. The current local configuration targets 0.4 megapixels for ordinary T2V and I2V and 0.3 megapixels for R2V. These are configuration targets, not promises of a fixed width and height for every file.
I2V derives the composition from the uploaded image. T2V and R2V expose aspect choices. The exact downloaded output should be checked in your media player or editor. An upstream model's maximum specification does not mean that every hosted preset uses that maximum.
Choose dimensions for the job
- Decide whether the final placement is a landscape video, square post or portrait Reel.
- Prepare an image source in that shape when using I2V.
- Generate a short draft and inspect both dimensions and subject framing.
- Plan any resize or crop in the final edit rather than assuming the prompt controls export dimensions.
Avoid a misleading quality label
Do not describe a file as 1080p or 2K merely because those words appear in its prompt. Inspect the encoded file. JollyAI's local H3 workflow should not be confused with MiniMax's separate full 2K pipeline. Upscaling also differs from generating additional scene detail at a higher native resolution.
Practical framing example
For a vertical product shot, begin with a portrait image showing the whole item and space for captions. For a website hero, choose landscape framing and keep the key action near the centre so responsive crops remain useful. Preserve the original download before making delivery-specific crops.
REVIEWED JOLLYAI IMPLEMENTATION
The workflow behind the guide
Written scene → text conditioning → turbo audio/video sampling → separate video and audio decoding → combined video file.
- Inputs, presets & access
- Limited free access; four-second free preset, with six/eight/ten-second PRO options. Current ordinary pixel budget: 0.4 MP. Use the account controls for current availability.
- Implementation note
- Local H3 turbo workflow. Six sampling steps; simple scheduler; turbo adapter; video/audio shifts 12/3. Current environment: pruned W4A8 FL2VA weights with a 4B encoder and matching projection. R2V uses a separate reference checkpoint. These settings describe the reviewed JollyAI deployment, not every H3 implementation.
Configuration reviewed September 20, 2026. Your current playground controls and plans page determine availability. Prompt recipes above are suggestions, not guarantees or claimed test results.
Open Text to video ↗Sources and useful next steps
The implementation notes describe JollyAI's reviewed local integration. Upstream capabilities and licences may differ from the settings available here.
- MiniMax H3 upstream project and model information ↗
- MiniMax H3 tool overview on JollyAI
- Earlier H3 ComfyUI workflow snapshot and downloadable JSON — check the newer configuration notes above when reproducing it.
- JollyAI content policy and service terms
Found an outdated setting or a reproducible problem? Contact JollyAI with the workflow and job reference. Keep account tokens and private input files out of public reports.