AI image model comparison
Hy Image 3.5 Preview vs MAI-Image-2.6 Flash
Image quality, cost per image, and editing features compared side by side. Or skip the guesswork and generate with both.
Tencent
Hy Image 3.5 Preview
Hy Image 3.5 Preview is a unified image generation and editing model from Tencent.
- Per image
- Not available
- Output / 1M
- $1.60
- Image Elo
- Not available
Microsoft
MAI-Image-2.6 Flash
MAI-Image-2.6 Flash is the lower-latency, lower-cost member of the [MAI-Image-2.6](/microsoft/mai-image-2.6) family from Microsoft AI, built for latency-sensitive, high-throughput production image generation and editing at comparable quality to the precision tier.
- Per image
- Not available
- Output / 1M
- $19
- Image Elo
- 1306
At a glance
Quick verdict
Which model wins on the things that matter most when you're making images.
Cheaper per image
Not available
Per-image price is not available
Better image quality
Not available
Not enough arena data
Better at editing
Not available
Only one editing score is available
Newer release
Hy Image 3.5 Preview
Released October 5, 2026
Quality
Image quality rankings
Elo ratings from Design Arena, where people vote between two anonymous images made from the same prompt. Higher is better.
Image generation
Image editing
Graphic design
Logos
Pricing
Hy Image 3.5 Preview vs MAI-Image-2.6 Flash price per image
What each image actually costs, plus the underlying API rates. Cheaper option highlighted.
| Metric | ||
|---|---|---|
| Cost per imageEstimated, ~1 megapixel (1024×1024) | Not available | Not available |
| Image outputPer 1M image tokens | $1.60 | $19 |
| Tokens per imageHow many image tokens one ~1MP image uses | Not available | Not available |
| Prompt textPer 1M input tokens | Included | $1.75 |
What it costs in practice
Estimated totals at standard resolution.
Single image
1 image
Social media batch
50 images
Product catalog
1,000 images
Specs
Inputs, outputs, and limits
What each model accepts, what it gives back, and how long your prompt can be.
| Metric | ||
|---|---|---|
| Accepts | Image, Text | Image, Text |
| Produces | Image | Image |
| Max prompt lengthText plus reference images | 100K tokens | 4K tokens |
| Reasoning step | No | No |
| Release date | October 5, 2026 | September 4, 2026 |
Features
Supported features
Creative controls available for each model.
| Metric | ||
|---|---|---|
| Image editingAccepts reference or source images | Supported | Supported |
| Text alongside imagesCan explain or caption what it made | Not supported | Not supported |
| Thinks before drawingPlans the image with a reasoning step | Not supported | Not supported |
| Reproducible seedsSame seed, same image | Not supported | Not supported |
| Temperature controlTrade consistency for variety | Not supported | Supported |
FAQ
Frequently asked questions
Can Hy Image 3.5 Preview and MAI-Image-2.6 Flash edit existing images?
Yes. Both accept reference images, so you can upload a photo and ask for changes like a new background, different lighting, or added text.
Can I use Hy Image 3.5 Preview and MAI-Image-2.6 Flash at the same time?
Yes. Shortcut Chat sends one prompt to multiple image models at once, so you can see Hy Image 3.5 Preview and MAI-Image-2.6 Flash results side by side and keep the one you like.
Why choose? Generate with both.
Send one prompt to Hy Image 3.5 Preview and MAI-Image-2.6 Flash at the same time. See both images side by side and keep the one you like.