MiniMax H3 Image Highspeed: 20MP, $0.0219/Image

The short answer
MiniMax H3 Image Highspeed, the faster variant of MiniMax H3 Image, is now live for text to image with sizes up to 20 megapixels and optional prompt expansion. In the AdGraam Playground it costs about $0.0219 per 1024×1024 image and $0.053 per 2048×2048 image. It is also callable through API providers.
On this page (10 sections)
What is MiniMax H3 Image Highspeed?
MiniMax H3 Image Highspeed is the faster variant of MiniMax's H3 Image model, now live for text to image. It supports custom output sizes up to 20 megapixels and optional prompt expansion that turns short prompts into full image descriptions.
It runs in the AdGraam Playground and is callable through API providers. A third-party write-up from Omniakey says the H3 Image family uses one 33.1B-parameter checkpoint for both generation and reference editing, with weights on Hugging Face under a custom community license. It lists two IDs, MiniMax-H3-Image and MiniMax-H3-Image-Highspeed, and says they share parameters and limits. Their official price examples differ.
It is separate from MiniMax H3, the video model. That video model is listed at $0.08 per second of video on MiniMax, per LiteLLM's price file.
MiniMax H3 Image Highspeed release date and availability
The Highspeed text to image model became callable on a model API on November 11, 2026, and is live in the AdGraam Playground. The edit variant is live there too.
Omniakey describes the wider H3 Image launch as covering the MiniMax API, MiniMax Code and Hailuo AI. MiniMax's own pages are not quoted here, so treat that list as a third-party summary. We covered the edit side in MiniMax H3 Image Edit and the LoRA release in MiniMax-H3-Image 2-Step LoRA.
If you already run the standard H3 Image model, the Highspeed ID is the one to test when turnaround and cost per image matter more than anything else. The parameters match, so swapping IDs in an existing workflow should not require changing your prompts or settings.
MiniMax H3 Image Highspeed key features and specs
The model generates from text at sizes up to 20 megapixels, with 1 to 10 images per run and four output formats. Prompt expansion is optional, which helps when briefs are short and you want the model to fill in scene detail.
| Spec | Text to Image | Edit |
|---|---|---|
| Max output | 20 megapixels | 20 megapixels |
| Reference images | None | Up to 16 |
| Aspect options | Square HD, Square, 3:4, 9:16, 4:3, 16:9 | Same, plus Auto |
| Images per run | 1 to 10 | 1 to 10 |
| Formats | jpeg, png, webp, jpg | jpeg, png, webp, jpg |
| Prompt expansion | Supported | Not listed |
| Price, 1024×1024 | about $0.0219 | about $0.0219 |
| Price, 2048×2048 | about $0.053 | about $0.053 |
Edit keeps the first image's aspect ratio by default and can combine multiple references through text instructions. Omniakey's guide says masked edits and transparent backgrounds are not supported, so cut-out product shots need another model.
Omniakey also reports that the API accepts an exact width and height, with an area between 655,360 and 20,000,000 pixels and aspect ratios from 1:3 to 3:1. It lists 4096×4096 as the 1:1 4K size, and says the model card gives about 16 MP sizes for other ratios. Those figures come from the standard H3 Image documentation, so confirm them for the Highspeed ID before building around them.
On quality, MiniMax claims more natural detail and stronger instruction following than other open-weight models, per that same write-up. Those are vendor claims. No independent benchmark for the Highspeed variant has been published yet.
MiniMax H3 Image Highspeed pricing and access
In the AdGraam Playground, a 1024×1024 image costs about $0.0219 and a 2048×2048 image about $0.053. The same rates apply to text to image and edit.
Coins and two sizes of printed photos on a desk representing per-image pricing
Other providers price differently. Runware's model page shows $0.04219 for an example text to image run and $0.074259 for a 2K reference edit. See AdGraam pricing for credit packs.
| Run | Playground price | Runware example |
|---|---|---|
| Text to image, 1024×1024 | about $0.0219 | Not listed |
| Text to image, 2048×2048 | about $0.053 | Not listed |
| Text to image, example prompt | Not applicable | $0.04219 |
| 2K reference edit, example prompt | Not applicable | $0.074259 |
At the Playground rate, 100 square images at 1024×1024 come to roughly $2.19. A batch of 10 images from one brief costs about 22 cents at that size, so you can test many prompt directions before spending real money on a winner.
How to use MiniMax H3 Image Highspeed for static ads
Use it for fast static ad exploration: many prompt variants at low cost, then refine the winners. The size menu includes Portrait 9:16 and Square options, which cover common vertical and feed formats.
A wall of square and vertical ad mockups in a creative studio
- Concept sweeps: generate up to 10 images per run from one brief, with prompt expansion on for short prompts.
- Product in context: use the edit variant with up to 16 references to place a product in new scenes, with the product photo as the first image so its aspect ratio carries through.
- Formats: pick Portrait 9:16 for vertical placements and Square for feed.
- Chaining: pass winners into upscaling or video in Flow Builder.
Runware's sample prompts include a magazine cover with exact headline text and a historic storefront restoration built from two reference photos. They suggest text rendering and reference-faithful editing are design targets. Check headline accuracy on your own copy before shipping, since a sample page proves little about your product.
For a full static workflow, see how to make static ads with AI or the Creative Generator. Teams that want to rebuild a competitor's winning layout can pair generations with Copy Ads.
A practical routine: write three to five short briefs, run each with prompt expansion on, review the 10-image batches, then rerun the best two at 2048×2048. Only that last step costs the higher $0.053 rate.
MiniMax H3 Image Highspeed vs other models
Highspeed is the cheaper, faster ID in the H3 Image family, but the speed claim is not independently measured. MiniMax says 4K generation takes under 10 seconds, per Omniakey, without naming a baseline, GPU or batch size. Omniakey says it did not run a speed test either.
For editing comparisons, look at FLUX 3 Image Edit, which takes 10 references against H3 Image Highspeed Edit's 16. Run both on your own product shots, since neither maker's samples reflect your catalog.
What to watch next
Watch for independent arena rankings of the Highspeed ID against the standard one, and for whether MiniMax adds masked edits. Standard versus Highspeed pricing at 4K is also worth checking once more providers list it. MiniMax has not announced any of these.
Key takeaways
- MiniMax H3 Image Highspeed text to image is live in the AdGraam Playground and callable through API providers.
- It supports custom output sizes up to 20 megapixels and optional prompt expansion.
- AdGraam Playground pricing is about $0.0219 per 1024×1024 image and $0.053 per 2048×2048 image.
- You can generate 1 to 10 images per run in jpeg, png, webp or jpg.
- The Highspeed Edit variant takes up to 16 reference images at the same per-image rates.
- MiniMax says 4K generation takes under 10 seconds, per a third-party write-up of its announcement, with no stated baseline.
Compare 200+ AI video and image models side by side
Run Kling, Veo, Sora, Seedance, Flux, Ideogram and Recraft in one Playground, or turn any product URL into ready-to-launch ads.
Frequently asked
What is MiniMax H3 Image Highspeed?
It is the Highspeed variant of MiniMax's H3 Image model, used here for text to image. It supports custom sizes up to 20 megapixels and optional prompt expansion.
How much does MiniMax H3 Image Highspeed cost?
In the AdGraam Playground it is about $0.0219 per 1024×1024 image and $0.053 per 2048×2048 image. Runware's model page lists $0.04219 for an example text to image run, so rates vary by provider and size.
How do I access MiniMax H3 Image Highspeed?
Run it in the AdGraam Playground, where it is live for text to image and edit. It is also available through API providers.
What aspect ratios does it support?
The Playground offers Square HD, Square, Portrait 3:4, Portrait 9:16, Landscape 4:3 and Landscape 16:9. The edit variant also offers Auto.
Is it the same as MiniMax H3?
No. MiniMax H3 is the video model. H3 Image is the image family, and Highspeed is its faster variant.
Sources
- MiniMax H3 Image Highspeedrunware.ai
- MiniMax H3 API Pricing (Nov 2026): $0.08 per Secondmodels.litellm.ai
- MiniMax H3-Image: 4K Images, Open Weights and API Accessomniakey.com
More AI model news & creative guides

MiniMax H3 Image Highspeed Edit: 16 Refs, $0.0219
MiniMax H3 Image Highspeed Edit takes up to 16 reference images and outputs up to 20MP from $0.0219 per image. Get the spec table, ad workflow and access.

Midjourney V8.2 Style Transfer: 4 Outputs, $0.129
Midjourney V8.2 Style Transfer restyles an image and returns 4 results for $0.129 per run in AdGraam Playground. Get the spec table, ad workflow and access.

Bria Fibo 2 Text to Image: Specs & $0.04 Price
Bria Fibo 2 (Fibo 2.0) text to image runs at $0.04 per image in AdGraam Playground. Get the spec table, ad workflow breakdown and access details.
Best AI Ad Generators for D2C Brands in 2026
Ten AI ad generators compared for D2C and e-commerce brands: what each makes, real starting prices, free options, and which suits Indian brands paying in rupees.
Break Even ROAS for COD Orders: How RTO Changes the Math
How to work out break even ROAS when a share of COD orders comes back as RTO, with worked rupee examples, a prepaid and COD mix, and why Ads Manager ROAS overstates COD.
What Indian D2C Brands Run on Meta: 2,790 Ads Analyzed
We studied 2,790 Facebook and Instagram ads from 14 Indian brands and their 747 longest-running ads: formats, angles, offers, languages and CTAs that last.
Your move
See your product as a winning ad
Turn a product URL into static ads and UGC-style videos, rebuild the competitor ads that already work, and read your Meta data to know what to make next.
AdGraam News DeskAI video and image news
The AdGraam News Desk (part of AdGraam, also searched as Adgram or Adgram AI) tracks AI image and video model launches, leaks and rumors from labs, leakers and creators, and checks each claim against its sources. Unconfirmed stories say so, and each page is updated as its story develops.