Skip to content

NVIDIA · Video

Cosmos 3 Super

NVIDIA 🇺🇸Video
From NanoGPT's listing:
Image-to-video generation from a first-frame image.
Market median price
$0.05 /sec
Listings
4
Providers
2

Evidence and sources

🤗 nvidia/cosmos3-superReleased 31 May 2026Maker HQ: US (source)
Sources: LiteLLM data, NanoGPT, llminfo researchPrimary data from LiteLLM from NanoGPT: name, description from llminfo research: release date

Input & output

as stated by the model's providers
Text
Images
Images
Video

Listings

 NanoGPTnvidia/cosmos-3-super/text-to-image · text-to-imageSource: NanoGPTNot published by the provider$0.04—
 NanoGPTnvidia/cosmos-3-super/image-to-video · image-to-videoSource: NanoGPTNot published by the provider—$0.05
 fal.aifal_ai/nvidia/cosmos-3-super/text-to-image · text-to-imageSource: LiteLLM dataNot published by the provider$0.04—
 fal.aifal_ai/nvidia/cosmos-3-super/image-to-video · image-to-videoSource: LiteLLM dataNot published by the provider—$0.05

Benchmarks

Named effort levels run from lowest to highest. Select a result to see its source details.

Video generation

Physical AI Bench21 related benchmarks · 33 results
Source: Physical AI Bench data
Generation
Conditional Generation Blur SSIM
Conditional Generation Depth si-RMSE
Conditional Generation Edge F1
Conditional Generation Quality Score
Show 15 more resultsHide 15 results

Safety

Data credits: Physical AI Bench (SHI Labs), MIT (licence)

Also known as: fal_ai/nvidia/cosmos-3-super/image-to-video, fal_ai/nvidia/cosmos-3-super/text-to-image, nvidia/cosmos-3-super/image-to-video, nvidia/cosmos-3-super/text-to-image

Like this one. Tap or hover any dotted value, label or icon to see where it comes from and how it is measured. open details when tapped or hovered. How this data works