MiniMax H3 Image VRAM (2026): The Encoder Is the Big File
ComfyUI added MiniMax H3 Image on 10 November 2026, and GGUF files followed the same day. The model card lists a 33.1B image transformer, but that is not the number that sets your VRAM. We read every file size from the Hugging Face API on 12 November 2026.
Bottom Line
- The text encoder is the biggest file, not the image model. The smallest ComfyUI set is 29.4 GB: a 12.55 GB image model, a 15.69 GB text encoder and a 1.17 GB VAE.
- 24GB is the comfortable tier. Every single file of the small set fits on the card with room for 4K output.
- 16GB should work with offload, unproven. ComfyUI loads the encoder and the image model in turn. The GGUF card says Q4_K_M needs “16 GB VRAM”, but it counts the image model only. Nobody had published a VRAM reading by 12 November 2026.
- Read the license first. It excludes the United States, the EU, the UK and South Korea. People there need a separate license from MiniMax.
| Your card | MiniMax H3 Image? | Files to use |
|---|---|---|
| 12GB | Possible, slow | Q3_K_M GGUF (8.91 GB) + NVFP4 encoder with offload |
| 16GB | Likely, with offload | Q4_K_M GGUF (11.57 GB) or w4a8 (12.55 GB) + NVFP4 encoder (15.69 GB) |
| 24GB | Yes | Q5_K_M (14.08 GB) or int8 (20.98 GB) + NVFP4 encoder |
| 32GB+ | Yes | int8 image model + int8 encoder (49.29 GB set) |
Plan on 32GB of system RAM at minimum for the 29.4 GB set; 64GB if you also load the optional prompt expander.
What MiniMax H3 Image Is
From the MiniMax model card and the Comfy blog (10 November 2026):
| Part | Value |
|---|---|
| Image model (DiT) | 33.1B parameters, dense, 50 layers, BF16 weights |
| Text encoder | The full Qwen3-VL text encoder of MiniMax H3, unchanged |
| Sampling | 6 steps, guidance 1, no negative prompt (fixed by the checkpoint) |
| Output | Native up to 4096×4096 (1:1), 5472×3072 (16:9), 6272×2688 (21:9) |
| Default in ComfyUI | 2048×2048 for text-to-image |
| Editing | Same checkpoint; connect up to several reference images |
| Prompt length | Up to 4,096 tokens |
It is not the H3 video model. Both share one text encoder, so an existing H3 setup already has that file.
Every File Size
Read from the Hugging Face API on 12 November 2026. GB means 10^9 bytes.
Official MiniMax repo (BF16, unpruned): transformer 66.28 GB, text encoder 66.73 GB, VAE 2.34 GB. That is 135 GB before quantization.
Comfy-Org repo (pruned image model):
| File | Size (GB) |
|---|---|
| Image model, pruned BF16 | 40.23 |
| Image model, pruned int8 | 20.98 |
| Image model, pruned w4a8 | 12.55 |
| Text encoder (Qwen3-VL 32B), int8 | 27.14 |
| Text encoder (Qwen3-VL 32B), NVFP4 | 15.69 |
| Prompt expander (Qwen3.5 9B), w4a8 | 7.34 |
| VAE, fp16 | 1.17 |
Abiray GGUF repo (image model only, from the pruned BF16 file):
| Quant | Size (GB) | Card’s VRAM advice |
|---|---|---|
| Q2_K | 6.88 | 8-10 GB |
| Q3_K_M | 8.91 | 12 GB |
| Q4_K_M | 11.57 | 16 GB |
| Q5_K_M | 14.08 | 16-24 GB |
The Arithmetic Nobody Prints
The GGUF card’s “16 GB VRAM” is for the image model alone. A working setup adds the encoder and the VAE:
| Set | Image model | Encoder | VAE | Total (GB) |
|---|---|---|---|---|
| Q3_K_M GGUF | 8.91 | 15.69 | 1.17 | 25.77 |
| Q4_K_M GGUF | 11.57 | 15.69 | 1.17 | 28.43 |
| Comfy w4a8 | 12.55 | 15.69 | 1.17 | 29.41 |
| Comfy int8 | 20.98 | 27.14 | 1.17 | 49.29 |
At every matched precision, the encoder is larger than the 4-bit image model. On 16GB, the encoder sets the floor, the same pattern as the H3 video model.
One more unknown: pruning. Every ComfyUI and GGUF file comes from a pruned image model. The pruned BF16 file is 40.23 GB against 66.28 GB for MiniMax’s transformer. Nobody has published what the pruning costs in image quality.
How to Run It in ComfyUI
From the Comfy blog:
- Update ComfyUI to the latest version.
- Download three files: an image model, the Qwen3-VL text encoder and the VAE.
- Load the MiniMax H3 Image template from the Templates panel.
- Write a long, complete prompt. To edit, connect images to
image_1onward and connect the VAE.
On 16GB, start at the 2048×2048 default. The 4096×4096 output needs far more activation memory; test it after the default works.
The License Catch
The MiniMax-H3-Image Community License grants rights only in the “Applicable Territory”: worldwide, excluding the European Union, the United Kingdom, the Republic of Korea and the United States. Use outside that territory is “not authorized”. Inside it, a business with more than $1 million in annual revenue, or anyone offering a hosted service, needs prior written authorization from MiniMax.
That is why this page links no hardware to buy: much of this site’s audience cannot use the model under these terms.
Sources
- MiniMaxAI/MiniMax-H3-Image model card, LICENSE and file list (read 2026-11-12).
- Comfy blog, “MiniMax H3 Image is now available in ComfyUI”, Rob, 10 November 2026.
- Hugging Face API: Comfy-Org/MiniMax-H3-Image (created 2026-11-08) and Abiray/MiniMax-H3-Image-GGUF (created 2026-11-10) file sizes and card, read 2026-11-12.
See Also
- Can I run MiniMax H3 locally?: the audio-video model with the same encoder
- Run Qwen Image 2.1 locally: the 12GB image model
- Is 16GB of VRAM still enough in 2026?: the tier-level verdict
- Best local LLM for an RTX 3090: what else the 24GB card runs
- Best local LLM for an RTX 5090: the 32GB tier
Sizing a card for image and video models? The tested hardware list groups GPUs by VRAM tier.