• 16:9 1920 × 1080
  • 4K UHD 3840 × 2160 is 16:9
  • 21:9 really 64:27 (2.37:1)
  • 9:16 Reels · Shorts · TikTok
  • 4:5 1080 × 1350 portrait posts
  • 27″ 1440p 108.8 PPI
  • 48 dp 144 px at xxhdpi
  • 1366 × 768 ≈ 16:9
  • 16:10 11% taller than 16:9
  • 2.39:1 cinema scope
  • 32:9 5120 × 1440
  • 3:2 6000 × 4000 camera sensors
  1. Home
  2. AI Image & Video Sizes

AI Image & Video Sizes

The sizes and aspect ratios each AI image and video model can make, and the best setting for the size you need.

15 models, each checked against its maker’s own documentation on . See every model in one table.

Updated

Find the best size for each AI model

Enter whole pixels, from 1 to 100,000.

The blank PNG is for apps with no size setting, such as the Gemini app: add it to your chat and ask for your picture on it. It’s made in your browser.

15 models for 1080 × 1920

  1. GPT Image 2.6 OpenAI · image

    Set
    size: "1088x1920"
    You get
    1088 × 1920
    Then
    Trim 4 px from the left and right
  2. Nano Banana 2.1 Google · image

    Set
    aspect_ratio: "9:16", image_size: "2K"
    You get
    1536 × 2752
    Then
    Scale to 1080 × 1935 (×0.703), then trim 7 px from the top and 8 px from the bottom
  3. Midjourney V8.2 Midjourney · image

    Set
    --ar 9:16 --hd
    You get
    Size not published
    Then
    Scale to 1080 × 1920 if needed
  4. FLUX 3 Image Black Forest Labs · image

    Set
    width: 1088, height: 1920
    You get
    1088 × 1920
    Then
    Trim 4 px from the left and right
  5. FLUX.2 Black Forest Labs · image

    Set
    width: 1088, height: 1920
    You get
    1088 × 1920
    Then
    Trim 4 px from the left and right
  6. Stable Diffusion XL Stability AI · image

    Set
    width: 768, height: 1344
    You get
    768 × 1344
    Then
    Scale up to 1097 × 1920 (×1.43), then trim 8 px from the left and 9 px from the right

    Scaling up ×1.43 softens detail.

  7. Stable Image API (SD 3.5, Ultra) Stability AI · image

    Set
    aspect_ratio: "9:16"
    You get
    768 × 1344
    Then
    Scale up to 1097 × 1920 (×1.43), then trim 8 px from the left and 9 px from the right

    Scaling up ×1.43 softens detail.

  8. Ideogram 4.6 Ideogram · image

    Set
    size: "1088x1920"
    You get
    1088 × 1920
    Then
    Trim 4 px from the left and right
  9. Adobe Firefly (web app) Adobe · image

    Set
    Aspect ratio: 3:4
    You get
    Size not published
    Then
    Trim 25% of the width, from the sides, then scale to 1080 × 1920

    1080 × 1920 (9:16) is taller than any shape it makes, so a lot is trimmed.

  10. Gemini Omni Flash 1.2 Google · video

    Set
    aspect_ratio: "9:16", resolution: "1080p"
    You get
    Size not published
    Then
    Scale to 1080 × 1920 if needed
  11. Veo 3.1 Google · video

    Set
    aspectRatio: "9:16", resolution: "1080p"
    You get
    Size not published
    Then
    Scale to 1080 × 1920 if needed
  12. Kling 4.0 Kling AI · video

    Set
    settings.aspect_ratio: "9:16", settings.resolution: "1080p"
    You get
    Size not published
    Then
    Scale to 1080 × 1920 if needed
  13. Runway Gen-4.5 · Image to video Runway · video

    Set
    ratio: "720:1280"
    You get
    720 × 1280
    Then
    Scale up to 1080 × 1920 (×1.5)

    Scaling up ×1.5 softens detail.

  14. Runway Gen-4.5 · Text to video Runway · video

    Set
    ratio: "720:1280"
    You get
    720 × 1280
    Then
    Scale up to 1080 × 1920 (×1.5)

    Scaling up ×1.5 softens detail.

  15. Luma Ray 3.2 Luma AI · video

    Set
    aspect_ratio: "9:16", video.resolution: "1080p"
    You get
    Size not published
    Then
    Scale to 1080 × 1920 if needed
  16. Seedance 2.5 ByteDance · video

    Set
    ratio: "9:16", resolution: "1080p"
    You get
    1080 × 1920
    Then
    Nothing to do

Image models

up to 3:1

OpenAI GPT Image 2.6

Image · released

Any width and height in multiples of 16, from 3:1 to 1:3, up to 3,840 px a side, up to 8,294,400 pixels in all and at least 655,360.

Set it with size: "WIDTHxHEIGHT"

Best setting for common sizes

  • 1080 × 1920 size: "1088x1920" Trim 4 px from the left and right
  • 1080 × 1350 size: "1088x1360" Scale to 1080 × 1350 (×0.993)
  • 1920 × 1080 size: "1920x1088" Trim 4 px from the top and bottom
  • 1280 × 720 size: "1280x720" Nothing to do
  • 1080 × 1080 size: "1088x1088" Scale to 1080 × 1080 (×0.993)

ChatGPT: The ChatGPT app has an aspect-ratio picker, or you can ask for a shape in your prompt. OpenAI doesn’t publish the sizes it makes.

In the API it’s gpt-image-2.6-sunburst or gpt-image-2.6-flare, released on November 4, 2026. size: "auto" lets the model choose.

Older models: gpt-image-1.5, gpt-image-1-mini and chatgpt-image-latest shut down on December 1, 2026, and gpt-image-2 on February 10, 2027.

14 shapes

Google Nano Banana 2.1

Image · released

14 shapes, each at 1K, 2K and 4K: 1:1, 2:3, 3:2, 3:4, 4:3, 4:5, 5:4, 9:16, 16:9, 21:9, 1:4, 4:1, 1:8 and 8:1.

Set it with aspect_ratio: "SHAPE", image_size: "TIER"

Best setting for common sizes

  • 1080 × 1920 aspect_ratio: "9:16", image_size: "2K" Scale to 1080 × 1935 (×0.703), then trim 7 px from the top and 8 px from the bottom
  • 1080 × 1350 aspect_ratio: "4:5", image_size: "2K" Scale to 1088 × 1350 (×0.586), then trim 4 px from the left and right
  • 1920 × 1080 aspect_ratio: "16:9", image_size: "2K" Scale to 1935 × 1080 (×0.703), then trim 7 px from the left and 8 px from the right
  • 1280 × 720 aspect_ratio: "16:9", image_size: "1K" Scale to 1290 × 720 (×0.938), then trim 5 px from the left and right
  • 1080 × 1080 aspect_ratio: "1:1", image_size: "2K" Scale to 1080 × 1080 (×0.527)

Gemini app: The Gemini app has no size setting; Google’s own examples put the shape in the prompt (“…with 2:3 aspect ratio”). For an exact shape, add a blank image the shape you want and ask for your picture on it.

In the API it’s gemini-nano-banana-2.1, generally available since October 6, 2026. Write the size with a capital K, as in 2K: lowercase is rejected.

If you don’t set a shape, the model matches the size of an image you add, or otherwise makes a square.

up to 14:1

Midjourney Midjourney V8.2

Image · released

Standard

Any whole-number ratio up to 14:1 or 1:14.

Set it with --ar W:H

HD

Any whole-number ratio up to 4:1 or 1:4.

Set it with --ar W:H --hd

Best setting for common sizes

  • 1080 × 1920 --ar 9:16 --hd Scale to 1080 × 1920 if needed
  • 1080 × 1350 --ar 4:5 --hd Scale to 1080 × 1350 if needed
  • 1920 × 1080 --ar 16:9 --hd Scale to 1927 × 1080 (×0.662), then trim 3 px from the left and 4 px from the right
  • 1280 × 720 --ar 16:9 Scale to 1285 × 720 (×0.882), then trim 2 px from the left and 3 px from the right
  • 1080 × 1080 --ar 1:1 --hd Scale to 1080 × 1080 (×0.527)

Whole numbers only: --ar 139:100, not --ar 1.39:1. V8.2 has been the default since July 24, 2026.

Midjourney publishes sizes for these four shapes only, and says HD sizes in some shapes may vary slightly. The HD sizes are also what its upscalers make from a standard image.

up to 8:1

Black Forest Labs FLUX 3 Image

Image · released

Any width and height in multiples of 16, from 8:1 to 1:8, 256 to 6,256 px a side, up to 16,822,272 pixels in all.

Set it with width: WIDTH, height: HEIGHT

Best setting for common sizes

  • 1080 × 1920 width: 1088, height: 1920 Trim 4 px from the left and right
  • 1080 × 1350 width: 1088, height: 1360 Scale to 1080 × 1350 (×0.993)
  • 1920 × 1080 width: 1920, height: 1088 Trim 4 px from the top and bottom
  • 1280 × 720 width: 1280, height: 720 Nothing to do
  • 1080 × 1080 width: 1088, height: 1088 Scale to 1080 × 1080 (×0.993)

BFL says the sizes it delivers are rounded to multiples of 16, so the sizes here are too. With reference images the limit is 4 MP.

Or set aspect_ratio (21:9 to 9:21) with resolution 768sq, 1k, 1.5k, 2k or 4k; a width and height override both. Width and height arrived on October 22, 2026.

any shape

Black Forest Labs FLUX.2

Image · released

Any width and height in multiples of 16, at least 64 px a side, up to 4,194,304 pixels in all.

Set it with width: WIDTH, height: HEIGHT

Best setting for common sizes

  • 1080 × 1920 width: 1088, height: 1920 Trim 4 px from the left and right
  • 1080 × 1350 width: 1088, height: 1360 Scale to 1080 × 1350 (×0.993)
  • 1920 × 1080 width: 1920, height: 1088 Trim 4 px from the top and bottom
  • 1280 × 720 width: 1280, height: 720 Nothing to do
  • 1080 × 1080 width: 1088, height: 1088 Scale to 1080 × 1080 (×0.993)

Up to 4 MP, which BFL illustrates as 2048 × 2048. Covers FLUX.2 [pro] and [flex], released on November 25, 2025, and [max], on December 16, 2025.

9 shapes

Stability AI Stable Diffusion XL

Image · released

9 sizes: 1024 × 1024, 1152 × 896, 896 × 1152, 1216 × 832, 832 × 1216, 1344 × 768, 768 × 1344, 1536 × 640 and 640 × 1536.

Set it with width: WIDTH, height: HEIGHT

Best setting for common sizes

  • 1080 × 1920 width: 768, height: 1344 Scale up to 1097 × 1920 (×1.43), then trim 8 px from the left and 9 px from the right
  • 1080 × 1350 width: 896, height: 1152 Scale up to 1080 × 1389 (×1.21), then trim 19 px from the top and 20 px from the bottom
  • 1920 × 1080 width: 1344, height: 768 Scale up to 1920 × 1097 (×1.43), then trim 8 px from the top and 9 px from the bottom
  • 1280 × 720 width: 1344, height: 768 Scale to 1280 × 731 (×0.952), then trim 5 px from the top and 6 px from the bottom
  • 1080 × 1080 width: 1024, height: 1024 Scale up to 1080 × 1080 (×1.05)

Stability calls these the best sizes for SDXL 1.0, announced on July 26, 2023; other sizes may crop the subject unexpectedly.

9 shapes

Stability AI Stable Image API (SD 3.5, Ultra)

Image

9 shapes, each at 1 MP: 21:9, 16:9, 3:2, 5:4, 1:1, 4:5, 2:3, 9:16 and 9:21.

Set it with aspect_ratio: "SHAPE"

Best setting for common sizes

  • 1080 × 1920 aspect_ratio: "9:16" Scale up to 1097 × 1920 (×1.43), then trim 8 px from the left and 9 px from the right
  • 1080 × 1350 aspect_ratio: "4:5" Scale up to 1112 × 1350 (×1.24), then trim 16 px from the left and right
  • 1920 × 1080 aspect_ratio: "16:9" Scale up to 1920 × 1097 (×1.43), then trim 8 px from the top and 9 px from the bottom
  • 1280 × 720 aspect_ratio: "16:9" Scale to 1280 × 731 (×0.952), then trim 5 px from the top and 6 px from the bottom
  • 1080 × 1080 aspect_ratio: "1:1" Scale up to 1080 × 1080 (×1.05)

Text to image only, and the default is 1:1; there’s no width or height setting.

Stable Image Core takes the same shapes, but Stability says it makes 1.5 MP, so its sizes may be larger than these.

up to 6:1

Ideogram Ideogram 4.6

Image · released

Any width and height in multiples of 32, from 6:1 to 1:6, 256 to 8,192 px a side, up to 16,777,216 pixels in all.

Set it with size: "WIDTHxHEIGHT"

Best setting for common sizes

  • 1080 × 1920 size: "1088x1920" Trim 4 px from the left and right
  • 1080 × 1350 size: "1120x1408" Scale to 1080 × 1358 (×0.964), then trim 4 px from the top and bottom
  • 1920 × 1080 size: "1920x1088" Trim 4 px from the top and bottom
  • 1280 × 720 size: "1312x736" Scale to 1283 × 720 (×0.978), then trim 1 px from the left and 2 px from the right
  • 1080 × 1080 size: "1088x1088" Scale to 1080 × 1080 (×0.993)

size can also be auto (currently 2K), auto_2k, auto_4k or source, the size of your first image. Released on November 19, 2026.

At very_low quality the limit is 2048 × 2048 pixels in all, and 4608 px a side.

4 shapes

Adobe Adobe Firefly (web app)

Image

4 shapes: 1:1, 4:3, 3:4 and 16:9.

Set it with Aspect ratio: SHAPE

Best setting for common sizes

  • 1080 × 1920 Aspect ratio: 3:4 Trim 25% of the width, from the sides, then scale to 1080 × 1920
  • 1080 × 1350 Aspect ratio: 3:4 Trim 6.3% of the height, from the top and bottom, then scale to 1080 × 1350
  • 1920 × 1080 Aspect ratio: 16:9 Scale to 1920 × 1080 if needed
  • 1280 × 720 Aspect ratio: 16:9 Scale to 1280 × 720 if needed
  • 1080 × 1080 Aspect ratio: 1:1 Scale to 1080 × 1080 if needed

Vertical 9:16 is only for the Firefly Image 4 and Image 4 Ultra models. Adobe doesn’t publish the pixel sizes.

Video models

2 shapes

Google Gemini Omni Flash 1.2

Video · released

2 shapes, at 360p, 720p, 1080p and 4k: 16:9 and 9:16.

Set it with aspect_ratio: "SHAPE", resolution: "TIER"

Best setting for common sizes

  • 1080 × 1920 aspect_ratio: "9:16", resolution: "1080p" Scale to 1080 × 1920 if needed
  • 1080 × 1350 aspect_ratio: "9:16", resolution: "1080p" Trim 29.7% of the height, from the top and bottom, then scale to 1080 × 1350
  • 1920 × 1080 aspect_ratio: "16:9", resolution: "1080p" Scale to 1920 × 1080 if needed
  • 1280 × 720 aspect_ratio: "16:9", resolution: "720p" Scale to 1280 × 720 if needed
  • 1080 × 1080 aspect_ratio: "16:9", resolution: "1080p" Trim 43.8% of the width, from the sides, then scale to 1080 × 1080

Flow and the Gemini app: Also in Flow, and rolling out in the Gemini app to Google AI Ultra subscribers.

The model is gemini-omni-1.2-flash, set with response_format. Google says 1080p and 4K are upscaled. Clips run 3 to 15 seconds.

2 shapes

Google Veo 3.1

Video · released

2 shapes, at 720p and 1080p: 16:9 and 9:16.

Set it with aspectRatio: "SHAPE", resolution: "TIER"

Best setting for common sizes

  • 1080 × 1920 aspectRatio: "9:16", resolution: "1080p" Scale to 1080 × 1920 if needed
  • 1080 × 1350 aspectRatio: "9:16", resolution: "1080p" Trim 29.7% of the height, from the top and bottom, then scale to 1080 × 1350
  • 1920 × 1080 aspectRatio: "16:9", resolution: "1080p" Scale to 1920 × 1080 if needed
  • 1280 × 720 aspectRatio: "16:9", resolution: "720p" Scale to 1280 × 720 if needed
  • 1080 × 1080 aspectRatio: "16:9", resolution: "1080p" Trim 43.8% of the width, from the sides, then scale to 1080 × 1080

On Google Cloud (formerly Vertex AI) as veo-3.1-generate-001 and veo-3.1-fast-generate-001. Clips are 4, 6 or 8 seconds.

Google’s pages disagree about 4K: the model page lists it for veo-3.1-generate-001, but the API reference says 4K is for the preview models only. Check before you rely on it.

6 shapes

Kling AI Kling 4.0

Video · released

6 shapes, at 720p, 1080p and 4k: 21:9, 16:9, 9:16, 1:1, 4:3 and 3:4.

Set it with settings.aspect_ratio: "SHAPE", settings.resolution: "TIER"

Best setting for common sizes

  • 1080 × 1920 settings.aspect_ratio: "9:16", settings.resolution: "1080p" Scale to 1080 × 1920 if needed
  • 1080 × 1350 settings.aspect_ratio: "3:4", settings.resolution: "1080p" Trim 6.3% of the height, from the top and bottom, then scale to 1080 × 1350
  • 1920 × 1080 settings.aspect_ratio: "16:9", settings.resolution: "1080p" Scale to 1920 × 1080 if needed
  • 1280 × 720 settings.aspect_ratio: "16:9", settings.resolution: "720p" Scale to 1280 × 720 if needed
  • 1080 × 1080 settings.aspect_ratio: "1:1", settings.resolution: "1080p" Scale to 1080 × 1080 if needed

4K was added on November 19, 2026, and 4:3 and 3:4 on November 14. For image to video, the shape follows your image.

6 shapes

Runway Runway Gen-4.5

Video · released

Image to video

6 sizes: 1280 × 720, 720 × 1280, 1104 × 832, 832 × 1104, 960 × 960 and 1584 × 672.

Set it with ratio: "WIDTH:HEIGHT"

Text to video

2 sizes: 1280 × 720 and 720 × 1280.

Set it with ratio: "WIDTH:HEIGHT"

Best setting for common sizes

  • 1080 × 1920 · Image to video ratio: "720:1280" Scale up to 1080 × 1920 (×1.5)
  • 1080 × 1920 · Text to video ratio: "720:1280" Scale up to 1080 × 1920 (×1.5)
  • 1080 × 1350 · Image to video ratio: "832:1104" Scale up to 1080 × 1433 (×1.3), then trim 41 px from the top and 42 px from the bottom
  • 1080 × 1350 · Text to video ratio: "720:1280" Scale up to 1080 × 1920 (×1.5), then trim 285 px from the top and bottom
  • 1920 × 1080 · Image to video ratio: "1280:720" Scale up to 1920 × 1080 (×1.5)
  • 1920 × 1080 · Text to video ratio: "1280:720" Scale up to 1920 × 1080 (×1.5)
  • 1280 × 720 · Image to video ratio: "1280:720" Nothing to do
  • 1280 × 720 · Text to video ratio: "1280:720" Nothing to do
  • 1080 × 1080 · Image to video ratio: "960:960" Scale up to 1080 × 1080 (×1.13)
  • 1080 × 1080 · Text to video ratio: "1280:720" Scale up to 1920 × 1080 (×1.5), then trim 420 px from the left and right

The model is gen4.5, and Runway’s ratio is the output size in pixels. Runway announced it on December 1, 2025, and it came to the API on February 10, 2026.

6 shapes

Luma AI Luma Ray 3.2

Video · released

6 shapes, at 360p, 540p, 720p and 1080p: 9:16, 3:4, 1:1, 4:3, 16:9 and 21:9.

Set it with aspect_ratio: "SHAPE", video.resolution: "TIER"

Best setting for common sizes

  • 1080 × 1920 aspect_ratio: "9:16", video.resolution: "1080p" Scale to 1080 × 1920 if needed
  • 1080 × 1350 aspect_ratio: "3:4", video.resolution: "1080p" Trim 6.3% of the height, from the top and bottom, then scale to 1080 × 1350
  • 1920 × 1080 aspect_ratio: "16:9", video.resolution: "1080p" Scale to 1920 × 1080 if needed
  • 1280 × 720 aspect_ratio: "16:9", video.resolution: "720p" Scale to 1280 × 720 if needed
  • 1080 × 1080 aspect_ratio: "1:1", video.resolution: "1080p" Scale to 1080 × 1080 if needed

The model is ray-3.2, released on June 9, 2026. Leave out the shape and it picks one from your prompt.

6 shapes

ByteDance Seedance 2.5

Video · released

6 shapes, each at 480p, 720p, 1080p and 4k: 16:9, 4:3, 1:1, 3:4, 9:16 and 21:9.

Set it with ratio: "SHAPE", resolution: "TIER"

Best setting for common sizes

  • 1080 × 1920 ratio: "9:16", resolution: "1080p" Nothing to do
  • 1080 × 1350 ratio: "3:4", resolution: "1080p" Scale to 1080 × 1440 (×0.865), then trim 45 px from the top and bottom
  • 1920 × 1080 ratio: "16:9", resolution: "1080p" Nothing to do
  • 1280 × 720 ratio: "16:9", resolution: "720p" Nothing to do
  • 1080 × 1080 ratio: "1:1", resolution: "1080p" Scale to 1080 × 1080 (×0.75)

Announced on July 31, 2026. Seedance 2.5 flash makes 480p and 720p only. The shape defaults to adaptive, which follows an image you add.

Every model in one table

Each model links to its card above.

ModelMakesSizesSet it withChecked
GPT Image 2.6
OpenAI
Image

Any width and height in multiples of 16, from 3:1 to 1:3, up to 3,840 px a side, up to 8,294,400 pixels in all and at least 655,360.

size: "WIDTHxHEIGHT"

Nano Banana 2.1
Google
Image

14 shapes, each at 1K, 2K and 4K: 1:1, 2:3, 3:2, 3:4, 4:3, 4:5, 5:4, 9:16, 16:9, 21:9, 1:4, 4:1, 1:8 and 8:1.

aspect_ratio: "SHAPE", image_size: "TIER"

Midjourney V8.2
Midjourney
Image

Standard: Any whole-number ratio up to 14:1 or 1:14.

HD: Any whole-number ratio up to 4:1 or 1:4.

--ar W:H

--ar W:H --hd

FLUX 3 Image
Black Forest Labs
Image

Any width and height in multiples of 16, from 8:1 to 1:8, 256 to 6,256 px a side, up to 16,822,272 pixels in all.

width: WIDTH, height: HEIGHT

FLUX.2
Black Forest Labs
Image

Any width and height in multiples of 16, at least 64 px a side, up to 4,194,304 pixels in all.

width: WIDTH, height: HEIGHT

Stable Diffusion XL
Stability AI
Image

9 sizes: 1024 × 1024, 1152 × 896, 896 × 1152, 1216 × 832, 832 × 1216, 1344 × 768, 768 × 1344, 1536 × 640 and 640 × 1536.

width: WIDTH, height: HEIGHT

Stable Image API (SD 3.5, Ultra)
Stability AI
Image

9 shapes, each at 1 MP: 21:9, 16:9, 3:2, 5:4, 1:1, 4:5, 2:3, 9:16 and 9:21.

aspect_ratio: "SHAPE"

Ideogram 4.6
Ideogram
Image

Any width and height in multiples of 32, from 6:1 to 1:6, 256 to 8,192 px a side, up to 16,777,216 pixels in all.

size: "WIDTHxHEIGHT"

Adobe Firefly (web app)
Adobe
Image

4 shapes: 1:1, 4:3, 3:4 and 16:9.

Aspect ratio: SHAPE

Gemini Omni Flash 1.2
Google
Video

2 shapes, at 360p, 720p, 1080p and 4k: 16:9 and 9:16.

aspect_ratio: "SHAPE", resolution: "TIER"

Veo 3.1
Google
Video

2 shapes, at 720p and 1080p: 16:9 and 9:16.

aspectRatio: "SHAPE", resolution: "TIER"

Kling 4.0
Kling AI
Video

6 shapes, at 720p, 1080p and 4k: 21:9, 16:9, 9:16, 1:1, 4:3 and 3:4.

settings.aspect_ratio: "SHAPE", settings.resolution: "TIER"

Runway Gen-4.5
Runway
Video

Image to video: 6 sizes: 1280 × 720, 720 × 1280, 1104 × 832, 832 × 1104, 960 × 960 and 1584 × 672.

Text to video: 2 sizes: 1280 × 720 and 720 × 1280.

ratio: "WIDTH:HEIGHT"

ratio: "WIDTH:HEIGHT"

Luma Ray 3.2
Luma AI
Video

6 shapes, at 360p, 540p, 720p and 1080p: 9:16, 3:4, 1:1, 4:3, 16:9 and 21:9.

aspect_ratio: "SHAPE", video.resolution: "TIER"

Seedance 2.5
ByteDance
Video

6 shapes, each at 480p, 720p, 1080p and 4k: 16:9, 4:3, 1:1, 3:4, 9:16 and 21:9.

ratio: "SHAPE", resolution: "TIER"

How the best size is chosen

Each model makes only certain sizes, so the size you need usually isn’t one of them. For each model, the calculator picks the size that:

  1. needs the least trimming to reach your shape (anything within 1% of the closest counts as just as close),
  2. doesn’t need scaling up, because scaling up softens detail,
  3. has a published pixel size, rather than an estimate,
  4. and is the smallest of those, so you don’t pay or wait for pixels you’ll throw away.

To finish, scale the picture until it covers your size, then trim what’s left over, half from each side. For a 1080 × 1920 video frame, GPT Image 2.6 makes 1088 × 1920: trim 4 px from the left and right. Nano Banana 2.1 makes 1536 × 2752: scale to 1080 × 1935 (×0.703), then trim 7 px from the top and 8 px from the bottom.

When the size isn’t published

Adobe Firefly (web app), Gemini Omni Flash 1.2, Veo 3.1, Kling 4.0 and Luma Ray 3.2 publish their shapes and a resolution such as 1080p, but not the exact pixel size of each, and the same setting can give different sizes in different shapes. For these the calculator shows the setting and the shape, and gives any trimming as a share of the picture rather than guessing a size.

Apps with no size setting

The Gemini app has no size setting, so the shape goes in your prompt. For an exact shape there’s a better way: Google’s documentation says its image models keep the size of an image you add, so add a blank image in the shape you want to your chat and ask for your picture on it. Download a blank PNG at your size above; it’s made in your browser. ChatGPT has its own aspect-ratio picker.

Retired models

These have shut down, though older guides and AI answers still mention them:

  • Sora (OpenAI, video). Shut down: App and website: April 26, 2026; API: September 24, 2026. There is no direct replacement. Source.
  • gpt-image-1 (OpenAI, image). Shut down: API: October 23, 2026. Use GPT Image 2.6 instead. OpenAI names gpt-image-2.5-sunburst or gpt-image-2.5-flare as its replacement; 2.6 is newer. Source.
  • Veo in the Gemini API (Google, video). Shut down: Veo 3.0 and 2.0: June 30, 2026; Veo 3.1 previews: October 22, 2026. Use Gemini Omni Flash 1.2 instead. Veo 3.1 is still on Google Cloud. Source.

Sources

Every size above comes from the model maker’s own documentation, checked on . Makers change their models often, so this page is checked again regularly.

More tools